How To Extract Snps From Chromosome 1-22 In One Command From *.Bed, *.Bim And *.Fam Files

Question

[User 0422]

Hi,

I need to extract chr 1-22 from SNP array data which also contains mtDNA and Y SNP data. Plink does it chromosome by chromosome and in this case I need to extract one chromosome by another and then join it, but I need something by which I can do it in one command.

Alternatively, the way to remove mt and Y SNPs from the main file would also work equally.

Please help if somebody can write the commands.

Thanks

[Aaronquinlan]

Providing an example of the contents of your file will likely lead to a quick answer.

Answer 1

[Maxime Lamontagne]

awk '{ if ($1 == 1 || $1 == 22) print $2 }' file.bim > snp.txt

plink --bfile file --extract snp.txt --make-bed --out newfile

[Maxime Lamontagne]

Sorry, I misunderstood your question.

awk '{ if ($1 >= 1 && $1 <= 22) print $2 }' file.bim > snp.txt

[User 0422]

this gives me only 1 and 22 numbers of chromosomes while I need 1 to 22! thanks

[brentp]

and you can get all chroms like: awk '($1 ~ /\d+/)' file.bim > snp.txt

[brentp]

^all numbered chroms^

[User 0422]

great! many thanks

[1234Jc4321]

thx Baboune. It helped me too.

Answer 2

[Chris Miller]

Assuming it's a bed file, or some other file that has the chromosome name in the first column, it's as simple as doing

perl -ne 'print $_ if /^chr[12]?[0-9]/' yourfile >outfile

As Aaron said, it's difficult to tell you about the others without a sample of the file format.

You really should read up on regular expressions, and the use of grep, awk, sed, or perl to do pattern matching.