Count number of SNPs per chromosome in vcf file

Question

[mostafarafiepour]

Hi all,

I want to count the number of SNP for each chromosome in the raw VCF file. What is the best idea?

Best Regard

Mostafa

[cpad0112]

Normalize your VCF and then execute: Datamash is in most of the linux repos

$ grep -v '^#' test.vcf | datamash -sg 1 count 1

with awk:

$ awk '!/^#/ { a[$1]++ } END {for (i in a) print i,a[i]}' test.vcf

[WouterDeCoster]

I have adapted your title to make it more descriptive of what you are asking.

Answer 1

[Pierre Lindenbaum]

UPDATE 2021: if your VCF is indexed: bcftools index -s indexed.vcf.gz

grep -v "^#" in.vcf | cut -f 1 | sort | uniq -c

[mostafarafiepour]

many thanks for your reply,

Does -f 1 mean the number of chromosomes?

[Pierre Lindenbaum]

https://linux.die.net/man/1/cut

[mostafarafiepour]

I apologize but this script does not work?

[Pierre Lindenbaum]

https://meta.stackexchange.com/questions/147616

[Kevin Blighe]

Pierre's script works for me, Moustafa:

grep -v "^#" test.vcf | cut -f 1 | sort | uniq -c
  16011 1
   7308 10
   9565 11
   9149 12
   3311 13
   5881 14
   5360 15
   7016 16
   8611 17
   2896 18
   9895 19
  11621 2
   3881 20
   2472 21
   3881 22
   9215 3
   7464 4
   7805 5
  10110 6
   7991 7
   6023 8
   6898 9
     37 MT
   3218 X
     21 Y

Chromosome 1 has 16011 variants... chromosome 9 has 6898, et cetera.

Your input VCF should be properly formatted and also be uncompressed.

[mostafarafiepour]

Yes i understood. Thank you very much for describing you.

[Pierre Lindenbaum]

If an answer was helpful you should upvote it, if the answer resolved your question you should mark it as accepted.

[mostafarafiepour]

Now, if i want to count the number of SNPs for each Breed, what is the best idea? i have 5 breed in the my raw vcf.

[cpad0112]

try VCFstats from RTGtools. But that would be stats per sample, not per chromosome. If you want per chromosome, per sample, then you may have to write a script. mostafarafiepour