Vt

From Genome Analysis Wiki
Revision as of 16:17, 21 October 2013 by Atks (talk | contribs)
Jump to navigationJump to search

Introduction

vt is a tool set that calls, genotypes and filters short variants. It provides profiling of variants to aid in QC.


Location

Internal usage 

  binaries
 /net/fantasia/home/atks/programs/vt

 test data
 /net/fantasia/home/atks/programs/vt/test

 scripts
 /net/fantasia/home/atks/programs/vt/scripts

External usage 

 download from sourceforge/github

Common options patterns

   -i defines the input file and by default, this is a require parameter,
      however, you may set it as '-' to accept STDIN which by default is 
      assumed to be a non compressed format.   

   -o defines the out file which and has the STDOUT set as the default.
      You may modify the STDOUT to output the binary version of the format, 
      e.g. BCF. with the option -c

Major Workflows

Discovery

Discovery is performed at per sample level, the evidence sites lists for each sample are then merged and site discovery statistics are computed. The user then makes a decision on cut offs to make to create an initial candidate site list.

Generates site list with info fields E and N. 

vt discover -i NA12878.bam -o NA12878.sites.vcf -g hs37d5.fa

Normalize(including left aligning) variants. This is required as left alignment of insertions and/or deletions within a read is sometimes insufficient to ensure complete left alignment. 

vt normalize -i NA12878.bam -o NA12878.normalized.sites.vcf -g hs37d5.fa

Evidence site lists are combined across samples and split by sites to allow for parallelization. 

vt merge_and_split_sample_vcf -i NA12878.sites.vcf,NA12879.sites.vcf,NA12880.sites.vcf -l 5000

Discovery statistics are computed. These statistics will allow you to choose a suitable cut off for creating a suitable candidate site list. 

vt compute_discovery_stats -i 1-1000000.sites.vcf -o 1-1000000.annotated.sites.vcf

Merge site lists. 

vt merge -i 1000000.sites.vcf,2000000.sites.vcf,3000000.sites.vcf -o candidate.sites.vcf

Plot charts to help with candidate list selection criteria. 

vt plot_discovery -i candidate.sites.vcf


A calling pipeline implemented in a make file is available here.

Genotyping

Each individual is genotyped at a set of sites. 

vt genotype -i NA12878.bam -o NA12878.sites.vcf -g hs37d5.fa

Genotype sample VCFs are combined across samples and split by sites. 

vt merge_and_split_sample_vcf -i NA12878.sites.vcf,NA12879.sites.vcf,NA12880.sites.vcf -o 1-1000000.sites.vcf

Features are computed. 

vt compute_features -i 1-1000000.sites.vcf -o 1-1000000.annotated.sites.vcf

A genotyping pipeline implemented in a make file is available here.

Filtering

Requires a set of features AND an installed copy of SVMLight. 

vt filter NA12878.bam -i NA12878.sites.vcf -o NA12878.svm.sites.vcf --pos positive.sites.vcf --neg negative.sites.vcf

A filtering pipeline implemented in a make file is available here.

Generation

Discovery

Discovers variants from bams. 

  Options:
  -b,  --input-bam-file    : Input BAM file
  -o,  --output-vcf-file   : Output VCF file
  -v,  --variant-type      : Variant Types, takes on any combinations of 
                             the values snps,mnps,indels comma delimited 
                             [snps,mnps,indels]
  -q,  --q-cutoff          : BASE Cutoff, only bases with 
                             QUAL/BAQ >= baseq are considered [13]
  -m,  --mapq-cutoff       : MAPQ Cutoff, only alignments with 
                             map quality >= mapq are considered [20]
  -g,  --genome-fa-file    : Genome FASTA file
  -s,  --sample-id         : Sample ID
  
  Example:
  e.g. vt discover -b in.bam -o - -g ref.fa -v snps,indels -s HG0001
  e.g. bam mergeBam --in a.bam --in b.bam -o - |
       vt discover -b - -o out.sites.vcf -g ref.fa -v all -s HG0001 | 
       vt left_align -i - | vt merge_duplicate_variants

Genotyping

Genotypes variants for each sample. 

  Options:
  -b,  --input-bam-file        : Input BAM file
  -i,  --input-candidate-vcf   : Input Candidate VCF file
  -o,  --output-vcf-file       : Output VCF file
  -v,  --variant-type          : Variant Types, takes on any combinations 
                                 of the values snps,mnps,indels comma 
                                 delimited [snps,mnps,indels]
  -g,  --genome-fa-file        : Genome FASTA file
  -s,  --sample-id             : Sample ID
  
  Example:
  e.g. vt genotype -b in.bam -i candidate.sites.vcf -o - -g ref.fa -s HG0001

Annotation

Make Probes

Populates the info field with REFPROBE, ALTPROBE and PLEN tags for genotyping. 

  Options:
  -i,  --input-vcf <string>      : Input VCF file
  -o,  --output-vcf <string>     : Output VCF file [-]
  -g,  --genome-fa               : Genome FASTA file [/net/fantasia/home/atks/ref/genome/human.g1k.v37.fa]
  -f,  --flank-length <integer>  : Minimum Flank Length [20]

  Example:
  e.g. vt make_probes -i 8904indels.dups.genotypes.vcf -o probes.sites.vcf -g ref.fa

Compute Feature

Compute feature of variant. 

vt compute_feature -i mills.vcf

Compute Allele balance

Compute allele balance. Outputs allele balance, allele frequency, genotype frequency. 

vt compute_ab -i mills.vcf

Compute Allele Frequency

Compute allele frequency. Outputs allele frequency and genotype frequency. 

vt compute_af -i mills.vcf

Compute Inbreeding Coefficient

Compute inbreeding coefficient. Outputs inbreeding coefficient based on genotype likelihoods. 

vt compute_fic -i mills.vcf

Compute HWE

Compute Hardy-Weinberg equilibrium statistic. Outputs PHRED scaled HWE Test p-values for biallelic as well as multiallelic variants. 

vt compute_hwe -i mills.vcf

Compute Mendelian Error

Compute mendelian error statistics. Outputs allele frequency and genotype frequency. 

vt compute_mendel -i mills.vcf

Compute features

vt compute_<feature1>_<feature2>_ ... _<feature n> -i mills.vcf

Modification

Left Alignment

Left aligns indel type variants in a VCF file. This differs from normalization in that it only left aligns and left trims a variant. This affects Indels only. 

vt left_align -i mills.vcf -o mills.leftaligned.vcf

Normalization

Normalize variants in a VCF file. 

vt normalize -i mills.vcf -o mills.normalized.vcf

Merge duplicate variants

Merges duplicate variants by position with the option of considering alleles. (This just discards the duplicate variant that appears later in the VCF file) 

  Options:
  -i,  --input-vcf <string>  : Input VCF file
  -o,  --output-vcf <string> : Output VCF file [-]
  -p,  --merge-by-position   : Merge by position [false]

  Example:
  e.g. vt merge_duplicate_variants -i 8904indels.dups.genotypes.vcf -o out.vcf
  e.g. vt merge_duplicate_variants -p -i 8904indels.dups.genotypes.vcf -o out.vcf

Profiling

A standard procedure is as follows: 

 zcat dataset.vcf.gx | vt normalize -i - | vt merge_duplicate_variants -i - > dataset.normalized.vcf

 cut -f1-8 dataset.normalized.vcf > dataset.sites.vcf

 cat dataset.normalized.sites.vcf | vt profile_snps -i - > snps.summary.log


Profile SNPs

Profile SNPs.

  • ts/tv ratio
  • overlap analyses
vt profile_snps -i mills.snps.sites.vcf

Profile Indels

Profile indels.

  • Overlap analyses with known data sets
  • FS/NFS annotation
vt profile_indels mills.indels.sites.vcf

Profile MNPs

Profile MNPs. 

vt profile_mnps -i mills.mnps.sites.vcf

Summarize Variants

Summarizes variants present in VCF file. 

vt peek -i mills.vcf

Plotting

Allele Frequency Spectrum

Plots Allele Frequency Spectrum of variants found in VCF file 

vt plot_afs -i mills.xml

Genotype Likelihood Concordance

Plots Genotype Likelihood Concordance graph. 

vt plot_gl -i mills.xml

Allele Balance Spectrum

Plots Allele Balance graph of variants in the VCF file. 

vt plot_ab -i mills.xml

VCF File Manipulation

Sort

Sort variants according to contig lists in header. 

vt sort -i mills.sites.vcf

Split by variant

Split VCF files by variant type. 

vt split_by_variant -i mills.sites.vcf

Resource Files

dbSNP OMNI 1000G Mills HAPMAP

Maintained by

This page is maintained by Adrian

Retrieved from "http://genome.sph.umich.edu/w/index.php?title=Vt&oldid=8869"