Changes

From Genome Analysis Wiki
Jump to navigationJump to search

SeqShop: Calling Your Own Genome, December 2014 (view source)

Revision as of 23:08, 8 December 2014

1,236 bytes removed ,  23:08, 8 December 2014
no edit summary
Line 37: Line 37:  
:<code>export OUT=~/$SAMPLE/output</code>
 
:<code>export OUT=~/$SAMPLE/output</code>
   −
 
+
== Step 1 (Day 1): Start SnpCall ==
−
== List of BAMs ==
+
=== List of BAMs ===
 
The list of BAMs has already been created (just 1 BAM, your sample).
 
The list of BAMs has already been created (just 1 BAM, your sample).
 
* But it is simply SAMPLE\tBAM_name, so easy to figure out
 
* But it is simply SAMPLE\tBAM_name, so easy to figure out
Line 47: Line 47:  
* Relative path, so assumes running from your home directory (I prefer absolute paths, but for simplicity of the workshop, we just use relative path).
 
* Relative path, so assumes running from your home directory (I prefer absolute paths, but for simplicity of the workshop, we just use relative path).
   −
== Configuring SnpCall ==
+
=== Configuring SnpCall ===
    
  cat ~/$SAMPLE/gotcloud.conf
 
  cat ~/$SAMPLE/gotcloud.conf
Line 73: Line 73:  
EXT_DIR = /net/seqshop-server/home/mktrost/seqshop/singleSample/ext
 
EXT_DIR = /net/seqshop-server/home/mktrost/seqshop/singleSample/ext
 
EXT = $(EXT_DIR)/ALL.chrCHR.phase3.combined.sites.unfiltered.vcf.gz $(EXT_DIR)/chrCHR.filtered.sites.vcf.gz
 
EXT = $(EXT_DIR)/ALL.chrCHR.phase3.combined.sites.unfiltered.vcf.gz $(EXT_DIR)/chrCHR.filtered.sites.vcf.gz
 +
</pre>
   −
</pre>
+
==== Configuration Updates ====
 +
In order to complete SnpCall overnight, we are going to tell GotCloud to only call SNPs for the EXOME regions.
 +
 
 +
Edit 
 +
nedit $SAMPLE/gotcloud.conf&
 +
* Or you can use <code>vi</code> or <code>emacs</code> or your favorite editor
 +
 
 +
Specify the target region:
 +
UNIFORM_TARGET_BED = /net/seqshop-server/home/mktrost/seqshop/singleSample/20130108.exome.targets.bed
 +
* See [[http://genome.sph.umich.edu/wiki/GotCloud:_Variant_Calling_Pipeline#Targeted.2FExome_Sequencing_Settings|Targeted/Exome Sequenceing Settings]] for more information on the GotCloud configuration settings for running Targeted/Exome runs.
   −
== Running SnpCall ==
+
=== Running SnpCall ===
 
Run GotCloud snpcall with 6 jobs running in parallel
 
Run GotCloud snpcall with 6 jobs running in parallel
 
* Why 6?   
 
* Why 6?   
Line 86: Line 96:  
This will run overnight.  We will check if it completed at the practical in the morning.
 
This will run overnight.  We will check if it completed at the practical in the morning.
   −
== Exome ==
+
=== Log Out ===
−
To speed things up, I extracted only exome regions from 100 1000g low coverage BAMs.
  −
 
  −
Let's create a new bam info file with your BAM combined with those BAMs.
  −
cp $ALIGN_OUT/bam.index $EXOME_OUT/bam.exome.index
  −
 
  −
Now add the exome BAMs to your new bam list:
  −
cat $IN/exome/bam.exome.index >> $EXOME_OUT/bam.exome.index
  −
 
  −
Verify you have 101 lines in your list:
  −
wc -l $EXOME_OUT/bam.exome.index
  −
 
  −
We are going to run on the cluster, so edit the first line of $EXOME_OUT/bam.exome.index to give the cluster path to your info file.
  −
nedit $EXOME_OUT/bam.exome.index
  −
Replace the /home on the first line with /net/seqshop-server
  −
 
  −
Locate your gotcloud.2x.conf (probably at: ~/personal/gotcloud.2x.conf) and open it in your favorite editor:
  −
nedit  ~/personal/gotcloud.2x.conf
  −
 
  −
Replace all occurrances of
  −
/home with /net/seqshop-server
  −
This is so you can run on the mini-cluster we have and can run more jobs at once
  −
 
  −
Update OUT_DIR & BAM_INDEX to:
  −
OUT_DIR = $(IN_DIR)/output.exome
  −
BAM_INDEX = $(OUT_DIR)/bam.exome.index
  −
 
  −
Update your gotcloud configuration file to indicate exomes:
  −
# Specify the path to the regions we want to call
  −
UNIFORM_TARGET_BED = $(REF_DIR)/20130108.exome.targets.nochr.bed
  −
  −
# We do not want any off target bases
  −
OFFSET_OFF_TARGET = 0
  −
  −
WRITE_TARGET_LOCI = TRUE
  −
TARGET_DIR = target
  −
 
  −
Remove CHRS = 20
  −
 
  −
Since it would take a while to run all 101 samples, I already ran the first step for the 100 1000G samples.
  −
We will "trick" GotCloud into thinking you already ran them by copying them into your output directory.
  −
cp -r $IN/exome/glfs $EXOME_OUT/.
  −
 
  −
 
  −
Run 4 jobs on our mini-cluster
  −
$GC/gotcloud snpcall --conf ~/personal/gotcloud.2x.conf --numjobs 4 --batchtype mosix --batchopts "-j10,11,12,13"
  −
* --batchtype says to use mosix (our cluster system)
  −
* --batchopts tells mosix the options to run with
  −
** for mosix, -j10,11,12,13 says to run on nodes 10, 11, 12, & 13 - the names of the 4 nodes on our mini-cluster
  −
 
  −
== Log Out ==
   
;Want to log out and leave your job running?
 
;Want to log out and leave your job running?
 
In the screen window, type:
 
In the screen window, type:
Line 148: Line 108:  
exit PuTTY
 
exit PuTTY
   −
== Day 2 (Tuesday) FEEDBACK! ==
+
=== Tuesday FEEDBACK! ===
 
Please provide feedback on the lectures/tutorials from today:  
 
Please provide feedback on the lectures/tutorials from today:  
    
https://docs.google.com/forms/d/1n8xYxvsOq-HsabpDfGcHvwD84BYIRDx8_b-H5N3d-D8/viewform
 
https://docs.google.com/forms/d/1n8xYxvsOq-HsabpDfGcHvwD84BYIRDx8_b-H5N3d-D8/viewform
   −
== Logging Back in to Check Jobs ==
+
== Step 2 (Day 2): Checking SnpCall ==
 +
=== Logging Back in to Check Jobs ===
    
;How do you log back into screen tomorrow?
 
;How do you log back into screen tomorrow?
 
  screen -r
 
  screen -r
 
This will resume an already running screen.
 
This will resume an already running screen.
Retrieved from "http://genome.sph.umich.edu/wiki/Special:MobileDiff/12224"

Navigation menu