Genome Analysis Wiki - User contributions [en]

Triodenovo

2023-01-11T22:19:02Z

Bingshan: /* Filtering */

== Note ==
The output genotypes for Indels are incorrect in this version but the de novo evidence is correctly calculated. The incorrect genotype "A/A" is the homozygous reference allele, "A/C" is the heterozygous, and "C/C" is the homozygous alternative allele, where the reference and alternative alleles for Indels are in the REF and ALT columns.

== Update ==
v0.06 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

https://github.com/yongzhuang/DNMFilter

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.06 [[Media:triodenovo.0.06.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

File:Triodenovo.0.06.tar.gz

2019-03-07T06:37:59Z

Bingshan:

Triodenovo

2019-03-07T06:37:37Z

Bingshan: /* Download */

== Note ==
The output genotypes for Indels are incorrect in this version but the de novo evidence is correctly calculated. The incorrect genotype "A/A" is the homozygous reference allele, "A/C" is the heterozygous, and "C/C" is the homozygous alternative allele, where the reference and alternative alleles for Indels are in the REF and ALT columns.

== Update ==
v0.06 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.06 [[Media:triodenovo.0.06.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Triodenovo

2019-03-07T06:37:28Z

Bingshan: /* Download */

== Note ==
The output genotypes for Indels are incorrect in this version but the de novo evidence is correctly calculated. The incorrect genotype "A/A" is the homozygous reference allele, "A/C" is the heterozygous, and "C/C" is the homozygous alternative allele, where the reference and alternative alleles for Indels are in the REF and ALT columns.

== Update ==
v0.06 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.06 [[Media:triodenovo.0.04.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Triodenovo

2019-03-07T06:37:08Z

Bingshan: /* Update */

== Note ==
The output genotypes for Indels are incorrect in this version but the de novo evidence is correctly calculated. The incorrect genotype "A/A" is the homozygous reference allele, "A/C" is the heterozygous, and "C/C" is the homozygous alternative allele, where the reference and alternative alleles for Indels are in the REF and ALT columns.

== Update ==
v0.06 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.04 [[Media:triodenovo.0.04.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

File:Triodenovo.0.04.tar.gz

2019-03-05T07:04:59Z

Bingshan: Bingshan uploaded a new version of File:Triodenovo.0.04.tar.gz

Triodenovo

2019-03-05T07:03:27Z

Bingshan: /* Download */

== Note ==
The output genotypes for Indels are incorrect in this version but the de novo evidence is correctly calculated. The incorrect genotype "A/A" is the homozygous reference allele, "A/C" is the heterozygous, and "C/C" is the homozygous alternative allele, where the reference and alternative alleles for Indels are in the REF and ALT columns.

== Update ==
v0.04 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.04 [[Media:triodenovo.0.04.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Triodenovo

2019-03-05T07:03:14Z

Bingshan: /* Update */

== Note ==
The output genotypes for Indels are incorrect in this version but the de novo evidence is correctly calculated. The incorrect genotype "A/A" is the homozygous reference allele, "A/C" is the heterozygous, and "C/C" is the homozygous alternative allele, where the reference and alternative alleles for Indels are in the REF and ALT columns.

== Update ==
v0.04 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.05 [[Media:triodenovo.0.05.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Polymutt2

2016-11-11T03:54:11Z

Bingshan: /* File format */

== Updates ==
The latest version of v0.2 is available for [[#Download | Download]].

== Note ==
'''Polymutt2 can only handle one chromosome, so please run it chromosome by chromosome'''.

== Compilation ==
*After downloading the source code, unzip and untar it, and cd polymutt2, and then type Make
*Two executables will be generated in bin/ directory: '''polymutt2''' and '''vcf2map'''
* vcf2map is to prune LD and generate a map file with high quality SNPs
* polymutt2 is to generate genotype calls taking a VCF file and a map file as input

== Usage ==
* First an LD pruned map file is generated by vcf2map. The following is the message by command vcf2map without any arguments

Input : --vcf [], --ped [],
--map [/scratch/cgg/lib13/db/hapmap/genetic_map_GRCh37_chr1.txt],
--include_list [/scratch/cgg/Public/hg19/1000G.SNV.clean.MAF0.05.tbl.gz]
Output : --out_map []
Variant filter : --min_maf [0.10], --min_avg_dp [0.00],
--max_avg_dp [-1.0e+00], --max_missing_rate [0.05]
LD pruning : --win_size [100], --max_r2 [0.10], --ignore_missing

* A command polymutt2 without any argument displays the following message

pedfile : (-pname)
datfile : (-dname)
mapfile : (-mname)

Additional Options
Input : --in_vcf [], --in_range [], --mixed_vcf_records
Mutation paramters : --theta_snv [1.0e-03], --theta_indel [1.0e-04],
--tstv_ratio [2.00], --submap [1.00]
Multi-threading : --nthreads [1]
Output : --out_vcf [], --fam_idx, --fam_id [], --out_all,
--out_range [], --best_marginal, --best_path
Approximation : --cum_prob [1.00], --single_iv

* NOTE: current version can only process one chromosome at a time

== Examples of generating the map file ==
* vcf2map: generate a sparse map file (see [[#Download|Download]] for files genetic_map_GRCh37_chr1.txt and 1000G.SNV.clean.MAF0.05.tbl.gz)
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --out_map chr1.map

* User defined r2 cutoff for LD pruning , min of average depth for filtering
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --max_r2 0.2 --min_avg_dp 2 --out_map chr1.r0.2.map

== Examples of running polymutt2 ==

* polymutt2: taking a VCF and the map file generated by vcf2map (the vcf file can be a complete vcf with all variants and samples)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf

* If parents are available genotypes can be phased by transmission (accuracy is not as good as above)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --best_path

* If a single family is desired to be output (the ped file can contain all families but will be ignored)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --fam_id

* If only a range is desire to output (for example the whole chromosome can be divided into multiple parallel jobs each working on a range)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --out_range 1:1000000-2000000

== File format ==
See PLINK http://pngu.mgh.harvard.edu/~purcell/plink/data.shtml for file format

== Download ==
*The latest version of source code v0.2 can be [[Media:Polymutt2_v0.2.tar.gz | downloaded]] here.
*The genetic map files (genetic_map_GRCh37_chr1.txt) used above can be [[Media:genetic_map_HapMapII_GRCh37.tar.gz | downloaded]] here.
*The clean and common variants in the 1000 Genome Project (1000G.SNV.clean.MAF0.05.tbl.gz) used above can be [[Media:1000G.SNV.clean.MAF0.05.tbl.gz | downloaded]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

== Citation ==
Li B, Wei Q, Zhan X, Zhong X, Chen W, Li C, et al. (2015) Leveraging Identity-by-Descent for Accurate Genotype Inference in Family Sequencing Data. PLoS Genet 11(6): e1005271. doi:10.1371/journal.pgen.1005271

Polymutt2

2016-11-11T03:52:14Z

Bingshan:

== Updates ==
The latest version of v0.2 is available for [[#Download | Download]].

== Note ==
'''Polymutt2 can only handle one chromosome, so please run it chromosome by chromosome'''.

== Compilation ==
*After downloading the source code, unzip and untar it, and cd polymutt2, and then type Make
*Two executables will be generated in bin/ directory: '''polymutt2''' and '''vcf2map'''
* vcf2map is to prune LD and generate a map file with high quality SNPs
* polymutt2 is to generate genotype calls taking a VCF file and a map file as input

== Usage ==
* First an LD pruned map file is generated by vcf2map. The following is the message by command vcf2map without any arguments

Input : --vcf [], --ped [],
--map [/scratch/cgg/lib13/db/hapmap/genetic_map_GRCh37_chr1.txt],
--include_list [/scratch/cgg/Public/hg19/1000G.SNV.clean.MAF0.05.tbl.gz]
Output : --out_map []
Variant filter : --min_maf [0.10], --min_avg_dp [0.00],
--max_avg_dp [-1.0e+00], --max_missing_rate [0.05]
LD pruning : --win_size [100], --max_r2 [0.10], --ignore_missing

* A command polymutt2 without any argument displays the following message

pedfile : (-pname)
datfile : (-dname)
mapfile : (-mname)

Additional Options
Input : --in_vcf [], --in_range [], --mixed_vcf_records
Mutation paramters : --theta_snv [1.0e-03], --theta_indel [1.0e-04],
--tstv_ratio [2.00], --submap [1.00]
Multi-threading : --nthreads [1]
Output : --out_vcf [], --fam_idx, --fam_id [], --out_all,
--out_range [], --best_marginal, --best_path
Approximation : --cum_prob [1.00], --single_iv

* NOTE: current version can only process one chromosome at a time

== Examples of generating the map file ==
* vcf2map: generate a sparse map file (see [[#Download|Download]] for files genetic_map_GRCh37_chr1.txt and 1000G.SNV.clean.MAF0.05.tbl.gz)
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --out_map chr1.map

* User defined r2 cutoff for LD pruning , min of average depth for filtering
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --max_r2 0.2 --min_avg_dp 2 --out_map chr1.r0.2.map

== Examples of running polymutt2 ==

* polymutt2: taking a VCF and the map file generated by vcf2map (the vcf file can be a complete vcf with all variants and samples)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf

* If parents are available genotypes can be phased by transmission (accuracy is not as good as above)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --best_path

* If a single family is desired to be output (the ped file can contain all families but will be ignored)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --fam_id

* If only a range is desire to output (for example the whole chromosome can be divided into multiple parallel jobs each working on a range)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --out_range 1:1000000-2000000

== File format ==
See http://genome.sph.umich.edu/wiki/Polymutt for file format

== Download ==
*The latest version of source code v0.2 can be [[Media:Polymutt2_v0.2.tar.gz | downloaded]] here.
*The genetic map files (genetic_map_GRCh37_chr1.txt) used above can be [[Media:genetic_map_HapMapII_GRCh37.tar.gz | downloaded]] here.
*The clean and common variants in the 1000 Genome Project (1000G.SNV.clean.MAF0.05.tbl.gz) used above can be [[Media:1000G.SNV.clean.MAF0.05.tbl.gz | downloaded]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

== Citation ==
Li B, Wei Q, Zhan X, Zhong X, Chen W, Li C, et al. (2015) Leveraging Identity-by-Descent for Accurate Genotype Inference in Family Sequencing Data. PLoS Genet 11(6): e1005271. doi:10.1371/journal.pgen.1005271

Bayesdenovo

2016-10-26T14:22:00Z

Bingshan: /* Filtering */

== Update ==
v0.01 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd bayesdenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''bayesdenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''nuclear families''' (including trios, quartets, and families with more siblings) for next-generation sequencing data. If infers Identity-by-Descednt (IBD) allele sharing to increase the '''''de novo''''' mutation calling accuracy. As a result, the IBD sharing for the called '''''de novo''''' mutations is also available in the output file.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message
*** This build (v.0.1) was compiled on Oct 25 2016, 11:16:22 ***
pedfile : (-pname)
datfile : (-dname)
mapfile : (-mname)

Additional Options
Input : --in_vcf [], --submap [1.00]
Denovo mutation parameters : --tstv_ratio [2.00], --minDQ [5.00]
Multi-threading : --nthreads [1]
Output : --out_prefix []

* Example 1: using default parameters

bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo --min_DQ 7

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows (Note that you can mix trios with other nuclear families in the same VCF file):
quartet1 p1 0 0 1
quartet1 p2 0 0 2
quartet1 p3 p1 p2 1
quartet1 p4 p1 p2 1
nuc1 p5 0 0 1
nuc1 p6 0 0 2
nuc1 p7 p1 p2 1
nuc1 p8 p1 p2 1
nuc1 p9 p1 p2 1
trio1 p10 0 0 1
trio1 p11 0 0 2
troi1 p12 p1 p2 1
trio2 p13 0 0 1
trio2 p14 0 0 2
troi2 p15 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

* A map file in the PLINK format. See blow for examples how to generate a map file with common and high quality variants

== Examples of generating the map file ==
* vcf2map: generate a sparse map file (see [[#Download|Download]] for files genetic_map_GRCh37_chr1.txt and 1000G.SNV.clean.MAF0.05.tbl.gz)
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --out_map chr1.map

* User defined r2 cutoff for LD pruning , min of average depth for filtering
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --max_r2 0.2 --min_avg_dp 2 --out_map chr1.r0.2.map

== Output ==
* The output will be one file per family, and the prefix to the names is specified via --out_prefix

An example of output file is as follows

##fileformat=VCFv4.1
##ProgramStart=Tue Oct 25 11:20:39 2016
##BayesDeNovo=../bin/bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo
##Note=VCF file modified by polymutt2. Updated fileds include: QUAL, GT and GQ, and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##INFO=<ID=FDQ,Number=1,Type=Integer,Description="Family-wise De Novo Mutatoin Quality in log10(BF) format">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Integer,Description="De Novo Mutation Quality in log10(BF) format">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=IV,Number=2,Type=Integer,Description="Best path inheritance vector. Founder alleles are arbiturally labeled (1 to 2*nFounders) and L1|L2 for non-founders indicated L1 and L2 from founders are transmitted">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT Rep1_1_1 Rep1_1_2 Rep1_1_3 Rep1_1_4
1 118 . C A 100 . AF=0.333333;AC=7;DP=1629;FDQ=8.667414 GT:DQ:DP:PL 0/0:.:202:1|2:0,100,255 0/0:.:234:3|4:0,100,255 0/1:8.97:203:1|3:100,0,255 0/0:.:206:2|4:0,100,255
1 858 . C A 100 . AF=0.333333;AC=10;DP=1592;FDQ=8.688325 GT:DQ:DP:PL 0/0:.:184:1|2:0,100,255 0/0:.:208:3|4:0,100,255 0/0:.:220:1|3:0,100,255 0/1:8.99:197:2|4:100,0,255

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advanced. Please see the triodenovo page below for more information:

http://genome.sph.umich.edu/wiki/Triodenovo

== Download ==
Source code of v0.01 [[Media:bayesdenovo.0.01.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Bayesdenovo

2016-10-26T14:21:23Z

Bingshan: /* Filtering */

== Update ==
v0.01 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd bayesdenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''bayesdenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''nuclear families''' (including trios, quartets, and families with more siblings) for next-generation sequencing data. If infers Identity-by-Descednt (IBD) allele sharing to increase the '''''de novo''''' mutation calling accuracy. As a result, the IBD sharing for the called '''''de novo''''' mutations is also available in the output file.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message
*** This build (v.0.1) was compiled on Oct 25 2016, 11:16:22 ***
pedfile : (-pname)
datfile : (-dname)
mapfile : (-mname)

Additional Options
Input : --in_vcf [], --submap [1.00]
Denovo mutation parameters : --tstv_ratio [2.00], --minDQ [5.00]
Multi-threading : --nthreads [1]
Output : --out_prefix []

* Example 1: using default parameters

bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo --min_DQ 7

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows (Note that you can mix trios with other nuclear families in the same VCF file):
quartet1 p1 0 0 1
quartet1 p2 0 0 2
quartet1 p3 p1 p2 1
quartet1 p4 p1 p2 1
nuc1 p5 0 0 1
nuc1 p6 0 0 2
nuc1 p7 p1 p2 1
nuc1 p8 p1 p2 1
nuc1 p9 p1 p2 1
trio1 p10 0 0 1
trio1 p11 0 0 2
troi1 p12 p1 p2 1
trio2 p13 0 0 1
trio2 p14 0 0 2
troi2 p15 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

* A map file in the PLINK format. See blow for examples how to generate a map file with common and high quality variants

== Examples of generating the map file ==
* vcf2map: generate a sparse map file (see [[#Download|Download]] for files genetic_map_GRCh37_chr1.txt and 1000G.SNV.clean.MAF0.05.tbl.gz)
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --out_map chr1.map

* User defined r2 cutoff for LD pruning , min of average depth for filtering
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --max_r2 0.2 --min_avg_dp 2 --out_map chr1.r0.2.map

== Output ==
* The output will be one file per family, and the prefix to the names is specified via --out_prefix

An example of output file is as follows

##fileformat=VCFv4.1
##ProgramStart=Tue Oct 25 11:20:39 2016
##BayesDeNovo=../bin/bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo
##Note=VCF file modified by polymutt2. Updated fileds include: QUAL, GT and GQ, and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##INFO=<ID=FDQ,Number=1,Type=Integer,Description="Family-wise De Novo Mutatoin Quality in log10(BF) format">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Integer,Description="De Novo Mutation Quality in log10(BF) format">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=IV,Number=2,Type=Integer,Description="Best path inheritance vector. Founder alleles are arbiturally labeled (1 to 2*nFounders) and L1|L2 for non-founders indicated L1 and L2 from founders are transmitted">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT Rep1_1_1 Rep1_1_2 Rep1_1_3 Rep1_1_4
1 118 . C A 100 . AF=0.333333;AC=7;DP=1629;FDQ=8.667414 GT:DQ:DP:PL 0/0:.:202:1|2:0,100,255 0/0:.:234:3|4:0,100,255 0/1:8.97:203:1|3:100,0,255 0/0:.:206:2|4:0,100,255
1 858 . C A 100 . AF=0.333333;AC=10;DP=1592;FDQ=8.688325 GT:DQ:DP:PL 0/0:.:184:1|2:0,100,255 0/0:.:208:3|4:0,100,255 0/0:.:220:1|3:0,100,255 0/1:8.99:197:2|4:100,0,255

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advanced. Please see the triodenovo page below for more information:

http://genome.sph.umich.edu/wiki/Triodenovo

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.01 [[Media:bayesdenovo.0.01.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Bayesdenovo

2016-10-25T19:51:29Z

Bingshan: /* Input files */

== Update ==
v0.01 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd bayesdenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''bayesdenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''nuclear families''' (including trios, quartets, and families with more siblings) for next-generation sequencing data. If infers Identity-by-Descednt (IBD) allele sharing to increase the '''''de novo''''' mutation calling accuracy. As a result, the IBD sharing for the called '''''de novo''''' mutations is also available in the output file.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message
*** This build (v.0.1) was compiled on Oct 25 2016, 11:16:22 ***
pedfile : (-pname)
datfile : (-dname)
mapfile : (-mname)

Additional Options
Input : --in_vcf [], --submap [1.00]
Denovo mutation parameters : --tstv_ratio [2.00], --minDQ [5.00]
Multi-threading : --nthreads [1]
Output : --out_prefix []

* Example 1: using default parameters

bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo --min_DQ 7

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows (Note that you can mix trios with other nuclear families in the same VCF file):
quartet1 p1 0 0 1
quartet1 p2 0 0 2
quartet1 p3 p1 p2 1
quartet1 p4 p1 p2 1
nuc1 p5 0 0 1
nuc1 p6 0 0 2
nuc1 p7 p1 p2 1
nuc1 p8 p1 p2 1
nuc1 p9 p1 p2 1
trio1 p10 0 0 1
trio1 p11 0 0 2
troi1 p12 p1 p2 1
trio2 p13 0 0 1
trio2 p14 0 0 2
troi2 p15 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

* A map file in the PLINK format. See blow for examples how to generate a map file with common and high quality variants

== Examples of generating the map file ==
* vcf2map: generate a sparse map file (see [[#Download|Download]] for files genetic_map_GRCh37_chr1.txt and 1000G.SNV.clean.MAF0.05.tbl.gz)
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --out_map chr1.map

* User defined r2 cutoff for LD pruning , min of average depth for filtering
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --max_r2 0.2 --min_avg_dp 2 --out_map chr1.r0.2.map

== Output ==
* The output will be one file per family, and the prefix to the names is specified via --out_prefix

An example of output file is as follows

##fileformat=VCFv4.1
##ProgramStart=Tue Oct 25 11:20:39 2016
##BayesDeNovo=../bin/bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo
##Note=VCF file modified by polymutt2. Updated fileds include: QUAL, GT and GQ, and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##INFO=<ID=FDQ,Number=1,Type=Integer,Description="Family-wise De Novo Mutatoin Quality in log10(BF) format">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Integer,Description="De Novo Mutation Quality in log10(BF) format">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=IV,Number=2,Type=Integer,Description="Best path inheritance vector. Founder alleles are arbiturally labeled (1 to 2*nFounders) and L1|L2 for non-founders indicated L1 and L2 from founders are transmitted">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT Rep1_1_1 Rep1_1_2 Rep1_1_3 Rep1_1_4
1 118 . C A 100 . AF=0.333333;AC=7;DP=1629;FDQ=8.667414 GT:DQ:DP:PL 0/0:.:202:1|2:0,100,255 0/0:.:234:3|4:0,100,255 0/1:8.97:203:1|3:100,0,255 0/0:.:206:2|4:0,100,255
1 858 . C A 100 . AF=0.333333;AC=10;DP=1592;FDQ=8.688325 GT:DQ:DP:PL 0/0:.:184:1|2:0,100,255 0/0:.:208:3|4:0,100,255 0/0:.:220:1|3:0,100,255 0/1:8.99:197:2|4:100,0,255

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance. Please see the triodenovo page below for more information:

http://genome.sph.umich.edu/wiki/Triodenovo

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.01 [[Media:bayesdenovo.0.01.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Bayesdenovo

2016-10-25T19:50:34Z

Bingshan: /* Input files */

== Update ==
v0.01 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd bayesdenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''bayesdenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''nuclear families''' (including trios, quartets, and families with more siblings) for next-generation sequencing data. If infers Identity-by-Descednt (IBD) allele sharing to increase the '''''de novo''''' mutation calling accuracy. As a result, the IBD sharing for the called '''''de novo''''' mutations is also available in the output file.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message
*** This build (v.0.1) was compiled on Oct 25 2016, 11:16:22 ***
pedfile : (-pname)
datfile : (-dname)
mapfile : (-mname)

Additional Options
Input : --in_vcf [], --submap [1.00]
Denovo mutation parameters : --tstv_ratio [2.00], --minDQ [5.00]
Multi-threading : --nthreads [1]
Output : --out_prefix []

* Example 1: using default parameters

bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo --min_DQ 7

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows (Note that you can mix trios with other nuclear families in the same VCF file):
quartet1 p1 0 0 1
quartet1 p2 0 0 2
quartet1 p3 p1 p2 1
quartet1 p4 p1 p2 1
nuc1 p5 0 0 1
nuc1 p6 0 0 2
nuc1 p7 p1 p2 1
nuc1 p8 p1 p2 1
nuc1 p9 p1 p2 1
trio1 p10 0 0 1
trio1 p11 0 0 2
troi1 p12 p1 p2 1
trio2 p13 0 0 1
trio2 p14 0 0 2
troi2 p15 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

* A map file in the PLINK format. See blow for examples how to generate a high quality map file.

== Examples of generating the map file ==
* vcf2map: generate a sparse map file (see [[#Download|Download]] for files genetic_map_GRCh37_chr1.txt and 1000G.SNV.clean.MAF0.05.tbl.gz)
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --out_map chr1.map

* User defined r2 cutoff for LD pruning , min of average depth for filtering
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --max_r2 0.2 --min_avg_dp 2 --out_map chr1.r0.2.map

== Output ==
* The output will be one file per family, and the prefix to the names is specified via --out_prefix

An example of output file is as follows

##fileformat=VCFv4.1
##ProgramStart=Tue Oct 25 11:20:39 2016
##BayesDeNovo=../bin/bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo
##Note=VCF file modified by polymutt2. Updated fileds include: QUAL, GT and GQ, and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##INFO=<ID=FDQ,Number=1,Type=Integer,Description="Family-wise De Novo Mutatoin Quality in log10(BF) format">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Integer,Description="De Novo Mutation Quality in log10(BF) format">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=IV,Number=2,Type=Integer,Description="Best path inheritance vector. Founder alleles are arbiturally labeled (1 to 2*nFounders) and L1|L2 for non-founders indicated L1 and L2 from founders are transmitted">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT Rep1_1_1 Rep1_1_2 Rep1_1_3 Rep1_1_4
1 118 . C A 100 . AF=0.333333;AC=7;DP=1629;FDQ=8.667414 GT:DQ:DP:PL 0/0:.:202:1|2:0,100,255 0/0:.:234:3|4:0,100,255 0/1:8.97:203:1|3:100,0,255 0/0:.:206:2|4:0,100,255
1 858 . C A 100 . AF=0.333333;AC=10;DP=1592;FDQ=8.688325 GT:DQ:DP:PL 0/0:.:184:1|2:0,100,255 0/0:.:208:3|4:0,100,255 0/0:.:220:1|3:0,100,255 0/1:8.99:197:2|4:100,0,255

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance. Please see the triodenovo page below for more information:

http://genome.sph.umich.edu/wiki/Triodenovo

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.01 [[Media:bayesdenovo.0.01.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Bayesdenovo

2016-10-25T16:38:30Z

Bingshan:

== Update ==
v0.01 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd bayesdenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''bayesdenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''nuclear families''' (including trios, quartets, and families with more siblings) for next-generation sequencing data. If infers Identity-by-Descednt (IBD) allele sharing to increase the '''''de novo''''' mutation calling accuracy. As a result, the IBD sharing for the called '''''de novo''''' mutations is also available in the output file.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message
*** This build (v.0.1) was compiled on Oct 25 2016, 11:16:22 ***
pedfile : (-pname)
datfile : (-dname)
mapfile : (-mname)

Additional Options
Input : --in_vcf [], --submap [1.00]
Denovo mutation parameters : --tstv_ratio [2.00], --minDQ [5.00]
Multi-threading : --nthreads [1]
Output : --out_prefix []

* Example 1: using default parameters

bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo --min_DQ 7

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
quartet1 p1 0 0 1
quartet1 p2 0 0 2
quartet1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

* A map file in the PLINK format. See blow for examples how to generate a high quality map file.

== Examples of generating the map file ==
* vcf2map: generate a sparse map file (see [[#Download|Download]] for files genetic_map_GRCh37_chr1.txt and 1000G.SNV.clean.MAF0.05.tbl.gz)
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --out_map chr1.map

* User defined r2 cutoff for LD pruning , min of average depth for filtering
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --max_r2 0.2 --min_avg_dp 2 --out_map chr1.r0.2.map

== Output ==
* The output will be one file per family, and the prefix to the names is specified via --out_prefix

An example of output file is as follows

##fileformat=VCFv4.1
##ProgramStart=Tue Oct 25 11:20:39 2016
##BayesDeNovo=../bin/bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo
##Note=VCF file modified by polymutt2. Updated fileds include: QUAL, GT and GQ, and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##INFO=<ID=FDQ,Number=1,Type=Integer,Description="Family-wise De Novo Mutatoin Quality in log10(BF) format">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Integer,Description="De Novo Mutation Quality in log10(BF) format">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=IV,Number=2,Type=Integer,Description="Best path inheritance vector. Founder alleles are arbiturally labeled (1 to 2*nFounders) and L1|L2 for non-founders indicated L1 and L2 from founders are transmitted">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT Rep1_1_1 Rep1_1_2 Rep1_1_3 Rep1_1_4
1 118 . C A 100 . AF=0.333333;AC=7;DP=1629;FDQ=8.667414 GT:DQ:DP:PL 0/0:.:202:1|2:0,100,255 0/0:.:234:3|4:0,100,255 0/1:8.97:203:1|3:100,0,255 0/0:.:206:2|4:0,100,255
1 858 . C A 100 . AF=0.333333;AC=10;DP=1592;FDQ=8.688325 GT:DQ:DP:PL 0/0:.:184:1|2:0,100,255 0/0:.:208:3|4:0,100,255 0/0:.:220:1|3:0,100,255 0/1:8.99:197:2|4:100,0,255

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance. Please see the triodenovo page below for more information:

http://genome.sph.umich.edu/wiki/Triodenovo

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.01 [[Media:bayesdenovo.0.01.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Bayesdenovo

2016-10-25T16:34:36Z

Bingshan: /* Introduction */

== Update ==
v0.01 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd bayesdenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''bayesdenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''nuclear families''' (including trios, quartets, and families with more siblings) for next-generation sequencing data. If infers Identity-by-Descednt (IBD) allele sharing to increase the '''''de novo''''' mutation calling accuracy. As a result, the IBD sharing for the called '''''de novo''''' mutations is also available in the output file.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message
*** This build (v.0.1) was compiled on Oct 25 2016, 11:16:22 ***
pedfile : (-pname)
datfile : (-dname)
mapfile : (-mname)

Additional Options
Input : --in_vcf [], --submap [1.00]
Denovo mutation parameters : --tstv_ratio [2.00], --minDQ [5.00]
Multi-threading : --nthreads [1]
Output : --out_prefix []

* Example 1: using default parameters

bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo --min_DQ 7

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
quartet1 p1 0 0 1
quartet1 p2 0 0 2
quartet1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output will be one file per family, and the prefix to the names is specified via --out_prefix

An example of output file is as follows

##fileformat=VCFv4.1
##ProgramStart=Tue Oct 25 11:20:39 2016
##BayesDeNovo=../bin/bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo
##Note=VCF file modified by polymutt2. Updated fileds include: QUAL, GT and GQ, and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##INFO=<ID=FDQ,Number=1,Type=Integer,Description="Family-wise De Novo Mutatoin Quality in log10(BF) format">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Integer,Description="De Novo Mutation Quality in log10(BF) format">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=IV,Number=2,Type=Integer,Description="Best path inheritance vector. Founder alleles are arbiturally labeled (1 to 2*nFounders) and L1|L2 for non-founders indicated L1 and L2 from founders are transmitted">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT Rep1_1_1 Rep1_1_2 Rep1_1_3 Rep1_1_4
1 118 . C A 100 . AF=0.333333;AC=7;DP=1629;FDQ=8.667414 GT:DQ:DP:PL 0/0:.:202:1|2:0,100,255 0/0:.:234:3|4:0,100,255 0/1:8.97:203:1|3:100,0,255 0/0:.:206:2|4:0,100,255
1 858 . C A 100 . AF=0.333333;AC=10;DP=1592;FDQ=8.688325 GT:DQ:DP:PL 0/0:.:184:1|2:0,100,255 0/0:.:208:3|4:0,100,255 0/0:.:220:1|3:0,100,255 0/1:8.99:197:2|4:100,0,255

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance. Please see the triodenovo page below for more information:

http://genome.sph.umich.edu/wiki/Triodenovo

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.01 [[Media:bayesdenovo.0.01.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

File:Bayesdenovo.0.01.tar.gz

2016-10-25T16:31:55Z

Bingshan:

Bayesdenovo

2016-10-25T16:30:31Z

Bingshan:

== Update ==
v0.01 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd bayesdenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''bayesdenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''nuclear families''' (including trios, quartets, and families with more siblings) for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message
*** This build (v.0.1) was compiled on Oct 25 2016, 11:16:22 ***
pedfile : (-pname)
datfile : (-dname)
mapfile : (-mname)

Additional Options
Input : --in_vcf [], --submap [1.00]
Denovo mutation parameters : --tstv_ratio [2.00], --minDQ [5.00]
Multi-threading : --nthreads [1]
Output : --out_prefix []

* Example 1: using default parameters

bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo --min_DQ 7

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
quartet1 p1 0 0 1
quartet1 p2 0 0 2
quartet1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output will be one file per family, and the prefix to the names is specified via --out_prefix

An example of output file is as follows

##fileformat=VCFv4.1
##ProgramStart=Tue Oct 25 11:20:39 2016
##BayesDeNovo=../bin/bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo
##Note=VCF file modified by polymutt2. Updated fileds include: QUAL, GT and GQ, and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##INFO=<ID=FDQ,Number=1,Type=Integer,Description="Family-wise De Novo Mutatoin Quality in log10(BF) format">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Integer,Description="De Novo Mutation Quality in log10(BF) format">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=IV,Number=2,Type=Integer,Description="Best path inheritance vector. Founder alleles are arbiturally labeled (1 to 2*nFounders) and L1|L2 for non-founders indicated L1 and L2 from founders are transmitted">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT Rep1_1_1 Rep1_1_2 Rep1_1_3 Rep1_1_4
1 118 . C A 100 . AF=0.333333;AC=7;DP=1629;FDQ=8.667414 GT:DQ:DP:PL 0/0:.:202:1|2:0,100,255 0/0:.:234:3|4:0,100,255 0/1:8.97:203:1|3:100,0,255 0/0:.:206:2|4:0,100,255
1 858 . C A 100 . AF=0.333333;AC=10;DP=1592;FDQ=8.688325 GT:DQ:DP:PL 0/0:.:184:1|2:0,100,255 0/0:.:208:3|4:0,100,255 0/0:.:220:1|3:0,100,255 0/1:8.99:197:2|4:100,0,255

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance. Please see the triodenovo page below for more information:

http://genome.sph.umich.edu/wiki/Triodenovo

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.01 [[Media:bayesdenovo.0.01.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Bayesdenovo

2016-10-25T16:25:59Z

Bingshan: Created page with "== Update == v0.01 is available for download == Compilation == * After downloading the source code, unzip and untar it, and cd bayesdenovo, and then type Make..."

== Update ==
v0.01 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd bayesdenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''bayesdenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''nuclear families''' (including trios, quartets, and families with more siblings) for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message
*** This build (v.0.1) was compiled on Oct 25 2016, 11:16:22 ***
pedfile : (-pname)
datfile : (-dname)
mapfile : (-mname)

Additional Options
Input : --in_vcf [], --submap [1.00]
Denovo mutation parameters : --tstv_ratio [2.00], --minDQ [5.00]
Multi-threading : --nthreads [1]
Output : --out_prefix []

* Example 1: using default parameters

bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo --min_DQ 7

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
quartet1 p1 0 0 1
quartet1 p2 0 0 2
quartet1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output will be one file per family, and the prefix to the names is specified via --out_prefix

An example of output file is as follows

##fileformat=VCFv4.1
##ProgramStart=Tue Oct 25 11:20:39 2016
##BayesDeNovo=../bin/bayesdenovo -p sim.vcf.ped -d sim.vcf.dat -m sim.vcf.map --in_vcf sim.vcf --out_prefix sim.vcf.denovo
##Note=VCF file modified by polymutt2. Updated fileds include: QUAL, GT and GQ, and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##INFO=<ID=FDQ,Number=1,Type=Integer,Description="Family-wise De Novo Mutatoin Quality in log10(BF) format">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Integer,Description="De Novo Mutation Quality in log10(BF) format">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=IV,Number=2,Type=Integer,Description="Best path inheritance vector. Founder alleles are arbiturally labeled (1 to 2*nFounders) and L1|L2 for non-founders indicated L1 and L2 from founders are transmitted">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT Rep1_1_1 Rep1_1_2 Rep1_1_3 Rep1_1_4
1 118 . C A 100 . AF=0.333333;AC=7;DP=1629;FDQ=8.667414 GT:DQ:DP:PL 0/0:.:202:1|2:0,100,255 0/0:.:234:3|4:0,100,255 0/1:8.97:203:1|3:100,0,255 0/0:.:206:2|4:0,100,255
1 858 . C A 100 . AF=0.333333;AC=10;DP=1592;FDQ=8.688325 GT:DQ:DP:PL 0/0:.:184:1|2:0,100,255 0/0:.:208:3|4:0,100,255 0/0:.:220:1|3:0,100,255 0/1:8.99:197:2|4:100,0,255

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.05 [[Media:triodenovo.0.05.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Idriver

2016-09-06T17:36:57Z

Bingshan: /* Introduction */

== Update ==
v0.1 is available for [[#Download | download]]

== Introduction ==

The software suite iDriver is to perform integrative modeling of multi-oimcs data to identify cancer driver genes. Details are in the READ file in the Download

== Download ==
Source code of v0.1 [[Media:iDriverV0.1.zip | download]] here.

== Contact ==
For questions please contact the authors (Hai Yang: [mailto:hai.yang@vanderbilt.edu hai.yang@vanderbilt.edu] or Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Idriver

2016-09-06T17:34:34Z

Bingshan:

== Update ==
v0.1 is available for [[#Download | download]]

== Introduction ==

The software suite iDriver is to perform integrative modeling of multi-oimcs data to identify cancer driver genes. Details are in the READ file in the [[#Download | download]]

== Download ==
Source code of v0.1 [[Media:iDriverV0.1.zip | download]] here.

== Contact ==
For questions please contact the authors (Hai Yang: [mailto:hai.yang@vanderbilt.edu hai.yang@vanderbilt.edu] or Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Idriver

2016-09-06T17:32:24Z

Bingshan:

== Update ==
v0.1 is available for [[#Download | download]]

== Introduction ==

The software suite iDriver is to perform integrative model of multi-oimcs data to identify cancer driver genes. Details are in the READ file in the [[#Download | download]]

== Download ==
Source code of v0.1 [[Media:iDriverV0.1.zip | download]] here.

== Contact ==
For questions please contact the authors (Hai Yang: [mailto:hai.yang@vanderbilt.edu hai.yang@vanderbilt.edu] or Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Idriver

2016-09-06T17:30:32Z

Bingshan:

== Download ==
Source code of v0.05 [[Media:iDriverV0.1.zip | download]] here.

== Contact ==
For questions please contact the authors (Hai Yang: [mailto:hai.yang@vanderbilt.edu hai.yang@vanderbilt.edu] or Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Idriver

2016-09-06T17:29:27Z

Bingshan:

== Download ==
Source code of v0.1 [[IDriverV0.1.zip | download]] here.

== Contact ==
For questions please contact the authors (Hai Yang: [mailto:hai.yang@vanderbilt.edu hai.yang@vanderbilt.edu] or Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Idriver

2016-09-06T17:28:40Z

Bingshan:

== Download ==
Source code of v0.1 [[iDriverV0.1.zip | download]] here.

== Contact ==
For questions please contact the authors (Hai Yang: [mailto:hai.yang@vanderbilt.edu hai.yang@vanderbilt.edu] or Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

File:IDriverV0.1.zip

2016-09-06T17:28:15Z

Bingshan:

Idriver

2016-09-06T17:26:51Z

Bingshan:

== Download ==
Source code of v0.1 [[iDriver0810.zip | download]] here.

== Contact ==
For questions please contact the authors (Hai Yang: [mailto:hai.yang@vanderbilt.edu hai.yang@vanderbilt.edu] or Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

File:IDriver 0810.zip

2016-09-06T17:26:09Z

Bingshan:

Idriver

2016-09-06T17:24:43Z

Bingshan: Created page with " == Download == Source code of v0.1 download here. == Contact == For questions please contact the authors (Hai Yang: [mailto:hai.yang@vanderbilt.edu h..."

== Download ==
Source code of v0.1 [[iDriver_0810.zip | download]] here.

== Contact ==
For questions please contact the authors (Hai Yang: [mailto:hai.yang@vanderbilt.edu hai.yang@vanderbilt.edu] or Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Triodenovo

2016-06-20T22:19:26Z

Bingshan: /* Usage */

== Note ==
The output genotypes for Indels are incorrect in this version but the de novo evidence is correctly calculated. The incorrect genotype "A/A" is the homozygous reference allele, "A/C" is the heterozygous, and "C/C" is the homozygous alternative allele, where the reference and alternative alleles for Indels are in the REF and ALT columns.

== Update ==
v0.05 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.05 [[Media:triodenovo.0.05.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Triodenovo

2016-06-20T22:11:52Z

Bingshan:

== Note ==
The output genotypes for Indels are incorrect in this version but the de novo evidence is correctly calculated. The incorrect genotype "A/A" is the homozygous reference allele, "A/C" is the heterozygous, and "C/C" is the homozygous alternative allele, where the reference and alternative alleles for Indels are in the REF and ALT columns.

== Update ==
v0.05 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

* Example 5: using --mixed_vcf_records when not all records have the same FORMAT. If NOT all FORMAT columns are the same, it will report error if this option is not provided.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --mixed_vcf_records

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.05 [[Media:triodenovo.0.05.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Triodenovo

2016-06-20T22:08:21Z

Bingshan:

== Update ==
v0.05 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

* Example 5: using --mixed_vcf_records when not all records have the same FORMAT. If NOT all FORMAT columns are the same, it will report error if this option is not provided.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --mixed_vcf_records

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.05 [[Media:triodenovo.0.05.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Triodenovo

2016-06-20T22:08:02Z

Bingshan: /* Update */

== Note ==

Recent VCF files tend to have records with different FORMAT. To avoid a potential problem caused by this please use '''--mixed_vcf_records''' on the command line when running triodenovo.

== Update ==
v0.05 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

* Example 5: using --mixed_vcf_records when not all records have the same FORMAT. If NOT all FORMAT columns are the same, it will report error if this option is not provided.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --mixed_vcf_records

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.05 [[Media:triodenovo.0.05.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

File:Triodenovo.0.05.tar.gz

2016-06-20T22:07:43Z

Bingshan:

Triodenovo

2016-06-20T22:07:27Z

Bingshan: /* Download */

== Note ==

Recent VCF files tend to have records with different FORMAT. To avoid a potential problem caused by this please use '''--mixed_vcf_records''' on the command line when running triodenovo.

== Update ==
v0.04 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

* Example 5: using --mixed_vcf_records when not all records have the same FORMAT. If NOT all FORMAT columns are the same, it will report error if this option is not provided.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --mixed_vcf_records

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.05 [[Media:triodenovo.0.05.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Triodenovo

2016-06-20T18:28:09Z

Bingshan: /* Note */

== Note ==

Recent VCF files tend to have records with different FORMAT. To avoid a potential problem caused by this please use '''--mixed_vcf_records''' on the command line when running triodenovo.

== Update ==
v0.04 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

* Example 5: using --mixed_vcf_records when not all records have the same FORMAT. If NOT all FORMAT columns are the same, it will report error if this option is not provided.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --mixed_vcf_records

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.04 [[Media:triodenovo.0.04.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Triodenovo

2016-06-20T18:27:28Z

Bingshan:

== Note ==

Recently VCF files tend to have records with different FORMAT. To avoid the problem please use '''--mixed_vcf_records''' on the command line when running triodenovo.

== Update ==
v0.04 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

* Example 5: using --mixed_vcf_records when not all records have the same FORMAT. If NOT all FORMAT columns are the same, it will report error if this option is not provided.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --mixed_vcf_records

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.04 [[Media:triodenovo.0.04.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Polymutt2

2016-04-07T16:15:33Z

Bingshan:

== Updates ==
The latest version of v0.2 is available for [[#Download | Download]].

== Note ==
'''Polymutt2 can only handle one chromosome, so please run it chromosome by chromosome'''.

== Compilation ==
*After downloading the source code, unzip and untar it, and cd polymutt2, and then type Make
*Two executables will be generated in bin/ directory: '''polymutt2''' and '''vcf2map'''
* vcf2map is to prune LD and generate a map file with high quality SNPs
* polymutt2 is to generate genotype calls taking a VCF file and a map file as input

== Usage ==
* First an LD pruned map file is generated by vcf2map. The following is the message by command vcf2map without any arguments

Input : --vcf [], --ped [],
--map [/scratch/cgg/lib13/db/hapmap/genetic_map_GRCh37_chr1.txt],
--include_list [/scratch/cgg/Public/hg19/1000G.SNV.clean.MAF0.05.tbl.gz]
Output : --out_map []
Variant filter : --min_maf [0.10], --min_avg_dp [0.00],
--max_avg_dp [-1.0e+00], --max_missing_rate [0.05]
LD pruning : --win_size [100], --max_r2 [0.10], --ignore_missing

* A command polymutt2 without any argument displays the following message

pedfile : (-pname)
datfile : (-dname)
mapfile : (-mname)

Additional Options
Input : --in_vcf [], --in_range [], --mixed_vcf_records
Mutation paramters : --theta_snv [1.0e-03], --theta_indel [1.0e-04],
--tstv_ratio [2.00], --submap [1.00]
Multi-threading : --nthreads [1]
Output : --out_vcf [], --fam_idx, --fam_id [], --out_all,
--out_range [], --best_marginal, --best_path
Approximation : --cum_prob [1.00], --single_iv

* NOTE: current version can only process one chromosome at a time

== Examples of generating the map file ==
* vcf2map: generate a sparse map file (see [[#Download|Download]] for files genetic_map_GRCh37_chr1.txt and 1000G.SNV.clean.MAF0.05.tbl.gz)
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --out_map chr1.map

* User defined r2 cutoff for LD pruning , min of average depth for filtering
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --max_r2 0.2 --min_avg_dp 2 --out_map chr1.r0.2.map

== Examples of running polymutt2 ==

* polymutt2: taking a VCF and the map file generated by vcf2map (the vcf file can be a complete vcf with all variants and samples)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf

* If parents are available genotypes can be phased by transmission (accuracy is not as good as above)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --best_path

* If a single family is desired to be output (the ped file can contain all families but will be ignored)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --fam_id

* If only a range is desire to output (for example the whole chromosome can be divided into multiple parallel jobs each working on a range)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --out_range 1:1000000-2000000

== Download ==
*The latest version of source code v0.2 can be [[Media:Polymutt2_v0.2.tar.gz | downloaded]] here.
*The genetic map files (genetic_map_GRCh37_chr1.txt) used above can be [[Media:genetic_map_HapMapII_GRCh37.tar.gz | downloaded]] here.
*The clean and common variants in the 1000 Genome Project (1000G.SNV.clean.MAF0.05.tbl.gz) used above can be [[Media:1000G.SNV.clean.MAF0.05.tbl.gz | downloaded]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

== Citation ==
Li B, Wei Q, Zhan X, Zhong X, Chen W, Li C, et al. (2015) Leveraging Identity-by-Descent for Accurate Genotype Inference in Family Sequencing Data. PLoS Genet 11(6): e1005271. doi:10.1371/journal.pgen.1005271

File:Genetic map HapMapII GRCh37.tar.gz

2016-02-15T19:57:46Z

Bingshan: Bingshan uploaded a new version of "File:Genetic map HapMapII GRCh37.tar.gz"

File:Genetic map HapMapII GRCh37.tar.gz

2016-02-15T19:57:16Z

Bingshan: Bingshan uploaded a new version of "File:Genetic map HapMapII GRCh37.tar.gz"

Triodenovo

2015-12-02T06:49:58Z

Bingshan: /* Usage */

== Update ==
v0.04 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

* Example 5: using --mixed_vcf_records when not all records have the same FORMAT. If NOT all FORMAT columns are the same, it will report error if this option is not provided.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --mixed_vcf_records

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.04 [[Media:triodenovo.0.04.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

File:Polymutt2 v0.2.tar.gz

2015-09-04T01:05:36Z

Bingshan:

Polymutt2

2015-09-04T01:04:31Z

Bingshan: /* Updates */

== Updates ==
The latest version of v0.2 is available for [[#Download | Download]].

== Compilation ==
*After downloading the source code, unzip and untar it, and cd polymutt2, and then type Make
*Two executables will be generated in bin/ directory: '''polymutt2''' and '''vcf2map'''
* vcf2map is to prune LD and generate a map file with high quality SNPs
* polymutt2 is to generate genotype calls taking a VCF file and a map file as input

== Usage ==
* First an LD pruned map file is generated by vcf2map. The following is the message by command vcf2map without any arguments

Input : --vcf [], --ped [],
--map [/scratch/cgg/lib13/db/hapmap/genetic_map_GRCh37_chr1.txt],
--include_list [/scratch/cgg/Public/hg19/1000G.SNV.clean.MAF0.05.tbl.gz]
Output : --out_map []
Variant filter : --min_maf [0.10], --min_avg_dp [0.00],
--max_avg_dp [-1.0e+00], --max_missing_rate [0.05]
LD pruning : --win_size [100], --max_r2 [0.10], --ignore_missing

* A command polymutt2 without any argument displays the following message

pedfile : (-pname)
datfile : (-dname)
mapfile : (-mname)

Additional Options
Input : --in_vcf [], --in_range [], --mixed_vcf_records
Mutation paramters : --theta_snv [1.0e-03], --theta_indel [1.0e-04],
--tstv_ratio [2.00], --submap [1.00]
Multi-threading : --nthreads [1]
Output : --out_vcf [], --fam_idx, --fam_id [], --out_all,
--out_range [], --best_marginal, --best_path
Approximation : --cum_prob [1.00], --single_iv

* NOTE: current version can only process one chromosome at a time

== Examples of generating the map file ==
* vcf2map: generate a sparse map file (see [[#Download|Download]] for files genetic_map_GRCh37_chr1.txt and 1000G.SNV.clean.MAF0.05.tbl.gz)
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --out_map chr1.map

* User defined r2 cutoff for LD pruning , min of average depth for filtering
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --max_r2 0.2 --min_avg_dp 2 --out_map chr1.r0.2.map

== Examples of running polymutt2 ==

* polymutt2: taking a VCF and the map file generated by vcf2map (the vcf file can be a complete vcf with all variants and samples)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf

* If parents are available genotypes can be phased by transmission (accuracy is not as good as above)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --best_path

* If a single family is desired to be output (the ped file can contain all families but will be ignored)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --fam_id

* If only a range is desire to output (for example the whole chromosome can be divided into multiple parallel jobs each working on a range)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --out_range 1:1000000-2000000

== Download ==
*The latest version of source code v0.2 can be [[Media:Polymutt2_v0.2.tar.gz | downloaded]] here.
*The genetic map files (genetic_map_GRCh37_chr1.txt) used above can be [[Media:genetic_map_HapMapII_GRCh37.tar.gz | downloaded]] here.
*The clean and common variants in the 1000 Genome Project (1000G.SNV.clean.MAF0.05.tbl.gz) used above can be [[Media:1000G.SNV.clean.MAF0.05.tbl.gz | downloaded]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

== Citation ==
Li B, Wei Q, Zhan X, Zhong X, Chen W, Li C, et al. (2015) Leveraging Identity-by-Descent for Accurate Genotype Inference in Family Sequencing Data. PLoS Genet 11(6): e1005271. doi:10.1371/journal.pgen.1005271

Polymutt2

2015-09-04T01:04:17Z

Bingshan: /* Download */

== Updates ==
The latest version of 0.1 is available for [[#Download | Download]].

== Compilation ==
*After downloading the source code, unzip and untar it, and cd polymutt2, and then type Make
*Two executables will be generated in bin/ directory: '''polymutt2''' and '''vcf2map'''
* vcf2map is to prune LD and generate a map file with high quality SNPs
* polymutt2 is to generate genotype calls taking a VCF file and a map file as input

== Usage ==
* First an LD pruned map file is generated by vcf2map. The following is the message by command vcf2map without any arguments

Input : --vcf [], --ped [],
--map [/scratch/cgg/lib13/db/hapmap/genetic_map_GRCh37_chr1.txt],
--include_list [/scratch/cgg/Public/hg19/1000G.SNV.clean.MAF0.05.tbl.gz]
Output : --out_map []
Variant filter : --min_maf [0.10], --min_avg_dp [0.00],
--max_avg_dp [-1.0e+00], --max_missing_rate [0.05]
LD pruning : --win_size [100], --max_r2 [0.10], --ignore_missing

* A command polymutt2 without any argument displays the following message

pedfile : (-pname)
datfile : (-dname)
mapfile : (-mname)

Additional Options
Input : --in_vcf [], --in_range [], --mixed_vcf_records
Mutation paramters : --theta_snv [1.0e-03], --theta_indel [1.0e-04],
--tstv_ratio [2.00], --submap [1.00]
Multi-threading : --nthreads [1]
Output : --out_vcf [], --fam_idx, --fam_id [], --out_all,
--out_range [], --best_marginal, --best_path
Approximation : --cum_prob [1.00], --single_iv

* NOTE: current version can only process one chromosome at a time

== Examples of generating the map file ==
* vcf2map: generate a sparse map file (see [[#Download|Download]] for files genetic_map_GRCh37_chr1.txt and 1000G.SNV.clean.MAF0.05.tbl.gz)
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --out_map chr1.map

* User defined r2 cutoff for LD pruning , min of average depth for filtering
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --max_r2 0.2 --min_avg_dp 2 --out_map chr1.r0.2.map

== Examples of running polymutt2 ==

* polymutt2: taking a VCF and the map file generated by vcf2map (the vcf file can be a complete vcf with all variants and samples)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf

* If parents are available genotypes can be phased by transmission (accuracy is not as good as above)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --best_path

* If a single family is desired to be output (the ped file can contain all families but will be ignored)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --fam_id

* If only a range is desire to output (for example the whole chromosome can be divided into multiple parallel jobs each working on a range)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --out_range 1:1000000-2000000

== Download ==
*The latest version of source code v0.2 can be [[Media:Polymutt2_v0.2.tar.gz | downloaded]] here.
*The genetic map files (genetic_map_GRCh37_chr1.txt) used above can be [[Media:genetic_map_HapMapII_GRCh37.tar.gz | downloaded]] here.
*The clean and common variants in the 1000 Genome Project (1000G.SNV.clean.MAF0.05.tbl.gz) used above can be [[Media:1000G.SNV.clean.MAF0.05.tbl.gz | downloaded]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

== Citation ==
Li B, Wei Q, Zhan X, Zhong X, Chen W, Li C, et al. (2015) Leveraging Identity-by-Descent for Accurate Genotype Inference in Family Sequencing Data. PLoS Genet 11(6): e1005271. doi:10.1371/journal.pgen.1005271

Polymutt2

2015-09-03T15:32:10Z

Bingshan: /* Contact */

== Updates ==
The latest version of 0.1 is available for [[#Download | Download]].

== Compilation ==
*After downloading the source code, unzip and untar it, and cd polymutt2, and then type Make
*Two executables will be generated in bin/ directory: '''polymutt2''' and '''vcf2map'''
* vcf2map is to prune LD and generate a map file with high quality SNPs
* polymutt2 is to generate genotype calls taking a VCF file and a map file as input

== Usage ==
* First an LD pruned map file is generated by vcf2map. The following is the message by command vcf2map without any arguments

Input : --vcf [], --ped [],
--map [/scratch/cgg/lib13/db/hapmap/genetic_map_GRCh37_chr1.txt],
--include_list [/scratch/cgg/Public/hg19/1000G.SNV.clean.MAF0.05.tbl.gz]
Output : --out_map []
Variant filter : --min_maf [0.10], --min_avg_dp [0.00],
--max_avg_dp [-1.0e+00], --max_missing_rate [0.05]
LD pruning : --win_size [100], --max_r2 [0.10], --ignore_missing

* A command polymutt2 without any argument displays the following message

pedfile : (-pname)
datfile : (-dname)
mapfile : (-mname)

Additional Options
Input : --in_vcf [], --in_range [], --mixed_vcf_records
Mutation paramters : --theta_snv [1.0e-03], --theta_indel [1.0e-04],
--tstv_ratio [2.00], --submap [1.00]
Multi-threading : --nthreads [1]
Output : --out_vcf [], --fam_idx, --fam_id [], --out_all,
--out_range [], --best_marginal, --best_path
Approximation : --cum_prob [1.00], --single_iv

* NOTE: current version can only process one chromosome at a time

== Examples of generating the map file ==
* vcf2map: generate a sparse map file (see [[#Download|Download]] for files genetic_map_GRCh37_chr1.txt and 1000G.SNV.clean.MAF0.05.tbl.gz)
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --out_map chr1.map

* User defined r2 cutoff for LD pruning , min of average depth for filtering
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --max_r2 0.2 --min_avg_dp 2 --out_map chr1.r0.2.map

== Examples of running polymutt2 ==

* polymutt2: taking a VCF and the map file generated by vcf2map (the vcf file can be a complete vcf with all variants and samples)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf

* If parents are available genotypes can be phased by transmission (accuracy is not as good as above)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --best_path

* If a single family is desired to be output (the ped file can contain all families but will be ignored)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --fam_id

* If only a range is desire to output (for example the whole chromosome can be divided into multiple parallel jobs each working on a range)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --out_range 1:1000000-2000000

== Download ==
*The latest version of source code v0.1 can be [[Media:Polymutt2.0.1.tar.gz | downloaded]] here.
*The genetic map files (genetic_map_GRCh37_chr1.txt) used above can be [[Media:genetic_map_HapMapII_GRCh37.tar.gz | downloaded]] here.
*The clean and common variants in the 1000 Genome Project (1000G.SNV.clean.MAF0.05.tbl.gz) used above can be [[Media:1000G.SNV.clean.MAF0.05.tbl.gz | downloaded]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

== Citation ==
Li B, Wei Q, Zhan X, Zhong X, Chen W, Li C, et al. (2015) Leveraging Identity-by-Descent for Accurate Genotype Inference in Family Sequencing Data. PLoS Genet 11(6): e1005271. doi:10.1371/journal.pgen.1005271

Triodenovo

2015-07-06T15:11:54Z

Bingshan: /* Input files */

== Update ==
v0.04 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.
** Note: In the VCF file either PL or GL has to be provided, and only the PL (or GL) field is used in the calling.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.04 [[Media:triodenovo.0.04.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Triodenovo

2015-07-06T15:09:34Z

Bingshan: /* Usage */

== Update ==
v0.04 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 3: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 4: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.04 [[Media:triodenovo.0.04.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Triodenovo

2015-07-06T15:08:57Z

Bingshan:

== Update ==
v0.04 is available for [[#Download | download]]

== Compilation ==
* After downloading the source code, unzip and untar it, and cd triodenovo, and then type Make
* If you encountered errors related to deprecated usage of some syntax please try to comment out the following in the core/Makefile
CXXFLAGS += -Werror -Wno-unused-variable -Wno-unused-result

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 1: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 1: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.04 [[Media:triodenovo.0.04.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Triodenovo

2015-07-06T15:06:19Z

Bingshan: /* Note */

== Update ==
v0.04 is available for [[#Download | download]]

== Introduction ==
* The program '''triodenovo''' implemented a Bayesian framework for calling '''''de novo''''' mutations in '''trios''' for next-generation sequencing data.
* It takes as input a standard VCF file with PL or GL fields (storing genotype likelihoods). Commonly used callers, e.g. GATK and samtools, generate VCF files with PL values.
* It calculates the likelihood of the model with '''''de novo''''' mutations, denoted as L1, and the likelihood of Mendelian transmission, denoted as L0, and represent the '''''de novo''''' evidence using a Bayesian factor BF=L1/L0. In TrioDeNovo the '''''de novo''''' quality is represented as DQ=log10(BF) = log10(L1/L0).
* DQ is the major parameter to control the output, along with others. See the [[#Output | example output ]] file below
* We recommend some basic and also a more advanced [[#Filtering | filtering]]

== Usage ==
A command without any input will invoke triodenovo and display the following message

The following parameters are in effect:
Input files : --ped [], --in_vcf []
Output files : --out_vcf []
Denovo mutation rate : --mu [1.0e-07]
Scaled mutation rate : --theta [1.0e-03], --indel_theta [1.0e-04]
Prior of de novo ts/tv ratio : --denovo_tstv [2.00]
Non-autosome labels : --chrX [X]
Filters : --minDQ [5.00], --minTotalDepth,
--maxTotalDepth, --minDepth [5], --maxDepth,
--mixed_vcf_records

* Example 1: using default parameters
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf

* Example 2: using --minDQ 7 to output de novo calls which are a minimum DQ of 7.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDQ 7

* Example 1: using --minDepth 10 to output de novo calls for which all three individuals (father, mother and child) have depth >=10.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minDepth 10

* Example 1: using --minTotalDepth 1000 to filter sites that have a total reads less than 1000. This option will filter the entire site.
triodenovo --ped trio.ped --in_vcf input.vcf --out denovo.vcf --minTotalDepth 1000

== Input files ==
* A ped file, with 5 colums [[http://www.sph.umich.edu/csg/abecasis/merlin/tour/ see merlin documentation]]. An example ped file is as follows
trio1 p1 0 0 1
trio1 p2 0 0 2
trio1 p3 p1 p2 1

* A VCF file [[http://www.1000genomes.org/node/101 VCF specs]]. It can contain variant information for more individuals than in the ped file.

== Output ==
* The output file is specified via --out_vcf

An example use is as follows

##fileformat=VCFv4.1
##triodenovo=../src/triodenovo --ped trio.denovo.ped --in_vcf trio.denovo.vcf --out_vcf trio.denovo.vcf.out
##Note=VCF file modified by polymutt. Updated fileds include: QUAL, GT and GQ, AF and AC. NOTE: modification was applied only to biallelic variants
##FILTER=<ID=LOWDP,Description="Low Depth filter when the average depth per sample is lessn than 1">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total Read Depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Alternative Allele Frequency">
##INFO=<ID=AC,Number=1,Type=Integer,Description="Alternative Allele Count">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=DQ,Number=1,Type=Denovo Quality,Description="Denovo Quality: log10(BF) where BF is Bayes Factor calculated as L(M1)/L(M0) for M1 and M0 representing models with and without de novo mutations">
##FORMAT=<ID=DGQ,Number=1,Type=Integer,Description="Denovo Genotype Quality: -10*log10(post) where post is the posterior probability of the called trio genotypes among all trios with de novo mutations">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
##FORMAT=<ID=PL,Number=3,Type=Integer,Description="Phred-scaled Genotype Likelihoods">
##FORMAT=<ID=GL,Number=3,Type=Float,Description="Log10 Genotype Likelihoods">
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT 3 2 1
1 10 . C T 20.71 . DP=47 GT:DQ:DGQ:DP:PL C/C:7.52:100:15:0,44,255 C/C:7.52:100:16:0,44,255 C/T:7.52:100:16:100,0,199
1 13 . C T 117.18 . DP=34 GT:DQ:DGQ:DP:PL C/C:5.67:100:8:0,23,197 C/C:5.67:100:11:0,32,255 C/T:5.67:100:15:178,0,103
1 17 . C A 111.15 . DP=36 GT:DQ:DGQ:DP:PL C/C:5.37:100:12:0,35,255 C/C:5.37:100:10:0,26,222 A/C:5.37:100:14:181,0,82
1 82 . C G 118.40 . DP=43 GT:DQ:DGQ:DP:PL C/C:6.45:100:14:0,41,255 C/C:6.45:100:13:0,38,255 C/G:6.45:100:16:199,0,100
1 91 . C T 28.69 . DP=36 GT:DQ:DGQ:DP:PL C/C:6.02:100:9:0,26,222 C/C:6.02:100:17:0,50,255 C/T:6.02:100:10:93,0,93

== Filtering ==
We recommend two filtering strategies. The first is a simple filtering and the second one is more advance

1. Basic filtering for SNVs. The following filter will retain sites of single nucleotides with only two alleles, QUAL>=30, and mutations in which parents are homozygous references and child is heterozygote with the heterozygote PL being zero, and the minimum PL of the other two genotypes in offering is 30 (i.e. the genotype likelihood, defined as P(R|G) in which R represents the aligned bases and G is the underlying genotype, of the called het mutation is >1000 than the genotype likelihood of the other two genotypes). These filtering parameters can be tuned as needed in the following command.

less trio.vcf.out | egrep "DQ|#" | perl -lane 'print if /#/; next if length($F[3])>1 || length($F[4])>1 || $F[4]=~/,/; next if $F[5]<30; $F[9] =~ /([A-Z])\/([A-Z])/; next if $1 ne $2; next if $F[10] !~ /$1\/$1/; $F[11]=~/([A-Z])\/([A-Z])/; next if $1 eq $2; $F[11] =~ /(\d+),(\d+),(\d+)/; next if $2 != 0 || $1<30 || $3<30; print' | less

2. Advanced filtering using a machine-learning approach (i.e. DNMFilter in the following webpage)

http://humangenome.duke.edu/software

3. Further thoughts about filtering for SNVs without bam files (step 2 requires bam files). There is no consensus on filtering so this can be very flexible.
* If you have a multi-sample call VCF it may be helpful to select those mutation candidates that appear only once in your VCF (AC=1 for example). This can be the top tier to consider. Relaxing AC to 2 or 3 can recover more real mutations but also increase false positives.
* If it is too stringent to filter out known sites, it may be helpful to select candidates that have low (e.g. <0.002)1000G or ESP allele frequencies. Some mutations can occur on know variant sites but mutations with high population frequencies may not be of great interest, if indeed they are real.
* Candidates in segmental duplications, low complexity regions or other copy number regions may be flagged for further analysis.
* Candidates for which parents are not hom-ref or offspring is a double mutant are more likely to be due to artifacts so the interpretation of these candidates may require additional QC if they appear to be interesting to the investigators.

== Download ==
Source code of v0.04 [[Media:triodenovo.0.04.tar.gz | download]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan.li@vanderbilt.edu bingshan.li@vanderbilt.edu])

Polymutt2

2015-07-02T03:34:35Z

Bingshan: /* Usage */

== Updates ==
The latest version of 0.1 is available for [[#Download | Download]].

== Compilation ==
*After downloading the source code, unzip and untar it, and cd polymutt2, and then type Make
*Two executables will be generated in bin/ directory: '''polymutt2''' and '''vcf2map'''
* vcf2map is to prune LD and generate a map file with high quality SNPs
* polymutt2 is to generate genotype calls taking a VCF file and a map file as input

== Usage ==
* First an LD pruned map file is generated by vcf2map. The following is the message by command vcf2map without any arguments

Input : --vcf [], --ped [],
--map [/scratch/cgg/lib13/db/hapmap/genetic_map_GRCh37_chr1.txt],
--include_list [/scratch/cgg/Public/hg19/1000G.SNV.clean.MAF0.05.tbl.gz]
Output : --out_map []
Variant filter : --min_maf [0.10], --min_avg_dp [0.00],
--max_avg_dp [-1.0e+00], --max_missing_rate [0.05]
LD pruning : --win_size [100], --max_r2 [0.10], --ignore_missing

* A command polymutt2 without any argument displays the following message

pedfile : (-pname)
datfile : (-dname)
mapfile : (-mname)

Additional Options
Input : --in_vcf [], --in_range [], --mixed_vcf_records
Mutation paramters : --theta_snv [1.0e-03], --theta_indel [1.0e-04],
--tstv_ratio [2.00], --submap [1.00]
Multi-threading : --nthreads [1]
Output : --out_vcf [], --fam_idx, --fam_id [], --out_all,
--out_range [], --best_marginal, --best_path
Approximation : --cum_prob [1.00], --single_iv

* NOTE: current version can only process one chromosome at a time

== Examples of generating the map file ==
* vcf2map: generate a sparse map file (see [[#Download|Download]] for files genetic_map_GRCh37_chr1.txt and 1000G.SNV.clean.MAF0.05.tbl.gz)
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --out_map chr1.map

* User defined r2 cutoff for LD pruning , min of average depth for filtering
vcf2map --vcf input.vcf --ped input.ped --map genetic_map_GRCh37_chr1.txt --include_list 1000G.SNV.clean.MAF0.05.tbl.gz --max_r2 0.2 --min_avg_dp 2 --out_map chr1.r0.2.map

== Examples of running polymutt2 ==

* polymutt2: taking a VCF and the map file generated by vcf2map (the vcf file can be a complete vcf with all variants and samples)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf

* If parents are available genotypes can be phased by transmission (accuracy is not as good as above)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --best_path

* If a single family is desired to be output (the ped file can contain all families but will be ignored)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --fam_id

* If only a range is desire to output (for example the whole chromosome can be divided into multiple parallel jobs each working on a range)
polymutt2 -p input.ped -m chr1.map --in_vcf input.vcf --out_vcf out.vcf --out_range 1:1000000-2000000

== Download ==
*The latest version of source code v0.1 can be [[Media:Polymutt2.0.1.tar.gz | downloaded]] here.
*The genetic map files (genetic_map_GRCh37_chr1.txt) used above can be [[Media:genetic_map_HapMapII_GRCh37.tar.gz | downloaded]] here.
*The clean and common variants in the 1000 Genome Project (1000G.SNV.clean.MAF0.05.tbl.gz) used above can be [[Media:1000G.SNV.clean.MAF0.05.tbl.gz | downloaded]] here.

== Contact ==
For questions please contact the authors (Bingshan Li: [mailto:bingshan@umich.edu bingshan@umich.edu])

== Citation ==
Li B, Wei Q, Zhan X, Zhong X, Chen W, Li C, et al. (2015) Leveraging Identity-by-Descent for Accurate Genotype Inference in Family Sequencing Data. PLoS Genet 11(6): e1005271. doi:10.1371/journal.pgen.1005271