SNP,INDEL counting per chromosomes in vcf

Question

[Siavash Salek Ardestani]

Hi all, how can I counting SNPs and INDELs per each chromosome in a vcf file?

Answer 1

[Alex Reynolds]

Using BEDOPS vcf2bed and standard Unix tools like wc to count lines:

$ vcf2bed --snvs < variants.vcf | wc -l
$ vcf2bed --insertions < variants.vcf | wc -l
$ vcf2bed --deletions < variants.vcf | wc -l

Cf. http://bedops.readthedocs.io/en/latest/content/reference/file-management/conversion/vcf2bed.html

The above does counts over all chromosomes. You need to do an extra step to count per chromosome.

To perform this count exercise quickly per-chromosome, per-variant class:

$ vcf2bed --snvs < variants.vcf > variants.snvs.bed
$ for chr in `bedextract --list-chr variants.snvs.bed`; do echo $chr; bedextract $chr variants.snvs.bed | wc -l; done

The bedextract tool is part of BEDOPS. If you install BEDOPS to get vcf2bed, you get bedextract, as well.

Repeat this procedure for --insertions and --deletions, replacing snvs accordingly, to count those classes.

[Siavash Salek Ardestani]

it is correct. I humbly thank you,

[Siavash Salek Ardestani]

I run these commands and I have number of SNP per chromosomes furthermore I have many 'Un' (I think this is unknown chromosomes SNP). how can I split or cut mt,x and unknown chromosomes from vcf?

[Alex Reynolds]

Use grep on the output of bedextract --list-chr. You can pass grep a file containing chromosome names you want to include (or, conversely, exclude) via grep -f.

$ for chr in `bedextract --list-chr variants.snvs.bed | grep -f wanted-chromosomes.txt -`; do ... ; done

Run man grep to see a full list of options.

[Siavash Salek Ardestani]

many thanks but I want cut mt, unknown and X chromosomes SNPs from a vcf file.

I means, my prior vcf have many SNPs that these SNPs are in mitochondrial DNA, X chromosome and unknown chromosomes. I want to have a vcf file without mitochondrial , X chromosome and unknown chromosomes SNPs information.

[Alex Reynolds]

You could use grep in the same way. It may help to review some examples of how it works so you can understand Unix streams and how grep can be used here.

[mostafarafiepour]

Hi, I also tried using this script but gave me only the number of snvs for each chromosome and did not count the insertions and deletions??

In fact, the outputs related to insertions and deletions are empty and there is no data in it?

[Ram]

Are you sure you have insertions/deletions?

[mostafarafiepour]

yes, Because when I use this command, i find 182456 deletions and 0 insertions.

insertions:

awk '! /\#/' variants.vcf | awk '{if(length($5) > 1) print}' | wc -l

insertions:

awk '! /\#/' variants.vcf | awk '{if(length($4) > 1 ) print}' | wc -l

Answer 2

[Pierre Lindenbaum]

using bioalcidaejdk: http://lindenb.github.io/jvarkit/BioAlcidaeJdk.html

$ java -jar dist/bioalcidaejdk.jar -e 'stream().map(V->V.getType().name()+" "+V.getContig()).collect(Collectors.groupingBy(Function.identity(), Collectors.counting())).forEach((K,V)->{println(K+" : "+V);});'  input.vcf

SNP 1 : 44
INDEL 1 : 6

[Siavash Salek Ardestani]

thanks but Did not respond

[mostafarafiepour]

Hi, i have been trying to do your scripts. But counted the only number of SNPs and did not count the InDels. ? Please see the photo in the attachment.

[Ram]

Read the part that begins with [WARN]

[Pierre Lindenbaum]

the warning is meaningless here.

[Ram]

Ah I see. It's probably my second guess then - no InDels to pick.

[Pierre Lindenbaum]

there is no INDEL in your VCF, in the htsjdk context : https://samtools.github.io/htsjdk/javadoc/htsjdk/htsjdk/variant/variantcontext/VariantContext.html#getType-- ; https://samtools.github.io/htsjdk/javadoc/htsjdk/htsjdk/variant/variantcontext/VariantContext.Type.html ;