This is a test version of Biostars. For the public version, visit https://www.biostars.org.
How to extract SNPs between samples and not between samples and reference

Hello,

I have used samtools mpileup, and bcftools call/filter (v1.5) to call SNPs between 2 samples and a de novo reference transcriptome.

Here is my pipeline:

bcftools mpileup -Ou -f ref.fa aln.bam1 aln.bam2 | \
    bcftools call -Ou -mv | \
    bcftools filter -s LowQual -e '%QUAL<20 || DP>100' > var.flt.vcf

I'm interested in extracting SNPs that occur between individual samples (aln.bam1 and aln.bam2) , and not between the samples and the reference (aln.bam 1 and ref.fa, aln.bam2 and ref.fa) . I would like to see how different each sample is different, or similar, from one another. Is this possible? I haven't been able to find much documentation on how one would approach this.

Any info would be greatly appreciated.

Cheers.

snp rna-seq
• 3,425 views • link
written by TrentGenomics • TrentGenomics

Thanks for your reply, Pierre. Currently running vcffilterjs now. Would the command be the same as the examples posted above, if there were more than 2 samples? For example, if there were 4 samples?

Thanks again.

• 0 views • link
written by TrentGenomics • TrentGenomics

of course not. But it's not clear from your question how you would filter out a VCF with N-Samples. However vcffilterjs can run some loops over the genotypes. see the example '''Sample having a unique genotype:'''

• 0 views • link
written by Pierre Lindenbaum Pierre Lindenbaum

My mistake, I'm quite new to bioinformatics. Thanks for your help so far. I've been running

 $ java -jar src/jvarkit-git/dist/vcffilterjs.jar -e 'variant.getGenotype(0).sameGenotype(variant.getGenotype(1))' input.var.flt.vcf

for ~24 hours now on a filtered .vcf file. What run times should I expect with the vcfffiltersjs program?

Thanks again

• 0 views • link
updated by Pierre Lindenbaum Pierre Lindenbaum • written by TrentGenomics • TrentGenomics

if the vcf input is huge, the processing is slow. you should see an estimation of the remaining time: something like:

[INFO][SAMSequenceDictionaryProgress]Count: 1994978 Elapsed: 3 hours(42.96%) Remains: 5 hours(57.04%) Last: 8:99026525
[INFO][SAMSequenceDictionaryProgress]Count: 2221249 Elapsed: 3 hours(49.37%) Remains: 4 hours(50.63%) Last: 9:138417471
[INFO][SAMSequenceDictionaryProgress]Count: 2200757 Elapsed: 3 hours(48.96%) Remains: 4 hours(51.04%) Last: 9:125920376
[INFO][SAMSequenceDictionaryProgress]Count: 2075688 Elapsed: 4 hours(45.47%) Remains: 4 hours(54.53%) Last: 9:17599369
[INFO][SAMSequenceDictionaryProgress]Count: 2051805 Elapsed: 3 hours(44.30%) Remains: 5 hours(55.70%) Last: 8:140497775
[INFO][SAMSequenceDictionaryProgress]Count: 1942671 Elapsed: 3 hours(41.35%) Remains: 5 hours(58.65%) Last: 8:48889823
[INFO][SAMSequenceDictionaryProgress]Count: 1996074 Elapsed: 4 hours(43.00%) Remains: 5 hours(57.00%) Last: 8:100040432
[INFO][SAMSequenceDictionaryProgress]Count: 2222632 Elapsed: 3 hours(49.39%) Remains: 4 hours(50.61%) Last: 9:139322248
[INFO][SAMSequenceDictionaryProgress]Count: 2202309 Elapsed: 3 hours(49.00%) Remains: 4 hours(51.00%) Last: 9:127025691

if not something is wrong.

Check your have really provided a input vcf, because vcfilterjs might try to wait for a VCF provided on standard input.

A faster alternative is my new tool: http://lindenb.github.io/jvarkit/VcfFilterJdk.html

 $ java -jar src/jvarkit-git/dist/vcffilterjdk.jar -e 'return variant.getGenotype(0).sameGenotype(variant.getGenotype(1));' input..vcf
• 0 views • link
written by Pierre Lindenbaum Pierre Lindenbaum

Thanks for getting back to me, this is a big help. The tool that you've copied above will return sample [0] not same genotype as sample [1]?

• 0 views • link
written by TrentGenomics • TrentGenomics

no the SAME genotype

for a different , that would be 

$ java -jar src/jvarkit-git/dist/vcffilterjdk.jar -e 'return !variant.getGenotype(0).sameGenotype(variant.getGenotype(1));' input..vcf

note the '!'

• 0 views • link
written by Pierre Lindenbaum Pierre Lindenbaum

Thank you so much, Pierre! I appreciate the help and insight.

• 0 views • link
written by TrentGenomics • TrentGenomics

Again, forgive me for my inexperience. I would like to output the results to a .txt file. Is this as simple as adding '-o somefile.txt' argument to the command? I tried to add an output argument but received an error:

java -jar /scratch/$USER/jvarkit/dist/vcffilterjdk.jar -e 'return !variant.getGenotype(0).sameGenotype(variant.getGenotype(1));' /scratch/$USER/SNPs/NFS_SFS_alignment.var.flt.vcf -o vcffilterjdk_different_genotype_NFS_SFS.txt

[SEVERE][Launcher]Must specify file or stream output type. java.lang.IllegalArgumentException: Must specify file or stream output type. at htsjdk.variant.variantcontext.writer.VariantContextWriterBuilder.build(VariantContextWriterBuilder.java:423) at com.github.lindenb.jvarkit.util.vcf.VCFUtils.createVariantContextWriter(VCFUtils.java:391) at com.github.lindenb.jvarkit.util.jcommander.Launcher.openVariantContextWriter(Launcher.java:946) at com.github.lindenb.jvarkit.util.jcommander.Launcher.doVcfToVcf(Launcher.java:981) at com.github.lindenb.jvarkit.util.jcommander.Launcher.doVcfToVcf(Launcher.java:1002) at com.github.lindenb.jvarkit.tools.vcffilterjs.VcfFilterJdk.doWork(VcfFilterJdk.java:322) at com.github.lindenb.jvarkit.util.jcommander.Launcher.instanceMain(Launcher.java:1155) at com.github.lindenb.jvarkit.util.jcommander.Launcher.instanceMainWithExit(Launcher.java:1341) at com.github.lindenb.jvarkit.tools.vcffilterjs.VcfFilterJdk.main(VcfFilterJdk.java:336) [INFO][Launcher]vcffilterjdk Exited with failure (-1)

Thanks again for guiding me through this.

• 0 views • link
written by TrentGenomics • TrentGenomics 
java -jar /scratch/$USER/jvarkit/dist/vcffilterjdk.jar -e 'return !variant.getGenotype(0).sameGenotype(variant.getGenotype(1));' /scratch/$USER/SNPs/NFS_SFS_alignment.var.flt.vcf > vcffilterjdk_different_genotype_NFS_SFS.vcf
• 0 views • link
written by Pierre Lindenbaum Pierre Lindenbaum

1 answer

I'm interested in extracting SNPs that occur between individual samples

using vcffilterjs: http://lindenb.github.io/jvarkit/VCFFilterJS.html

sample[0] same genotype as sample[1]

java -jar src/jvarkit-git/dist/vcffilterjs.jar -e 'variant.getGenotype(0).sameGenotype(variant.getGenotype(1))' input.vcf

sample[0] not same genotype as sample[1]

java -jar src/jvarkit-git/dist/vcffilterjs.jar -e '!variant.getGenotype(0).sameGenotype(variant.getGenotype(1))' input.vcf
• 0 views • link
written by Pierre Lindenbaum Pierre Lindenbaum

Log in to answer this question.

More posts like this