getting meaningful allele out of vcf file

[shawpa]

I have generated a VCF file from a kinome capture experiment after all the filtering steps. I have a 9.5Mb vcf file from GATK and I don't know what to do with it. Basically I would like to generate a file for each chr that has columns such as position, number of reads, how many times allele is called etc. I don't have any idea what program to use. I don't really care about annotating at this point, just want to see allele calls in a readable text file. Please help.

[pbluescript]

You should check out bedtools.

http://code.google.com/p/bedtools/

[shawpa]

I was looking into bedtools but I don't know which command to use.

[dpryan]

The VCF file is a text file with that information in it (largely separated into columns in fact). You can just open it in your favorite text document view/editor (MS Word, Notepad, gvim, Preview.app). Frankly, the easiest way to split the calls by chromosome is probably just to use grep (assuming you're using Linux or a Mac). So "grep chrX file.vcf > chrX.vcf" for each chromosome. I'm sure someone can think of a short one line command with awk and sed to avoid having to grep for each chromosome, but frankly you probably have a small number of chromosomes so this won't be very labor intensive. Note that the result of the grep command won't really be a valid VCF file, since you'll miss much of the header, but that won't matter if you just want to go through it by hand.

[shawpa]

The VCF file is a text file with that information in it (largely separated into columns in fact). You can just open it in your favorite text document view/editor (MS Word, Notepad, gvim, Preview.app). Frankly, the easiest way to split the calls by chromosome is probably just to use grep (assuming you're using Linux or a Mac). So "grep chrX file.vcf > chrX.vcf" for each chromosome. I'm sure someone can think of a short one line command with awk and sed to avoid having to grep for each chromosome, but frankly you probably have a small number of chromosomes so this won't be very labor intensive. Note that the result of the grep command won't really be a valid VCF file, since you'll miss much of the header, but that won't matter if you just want to go through it by hand.

Thanks! I get that this is probably a stupid question, but if I type grep chr1, it pulls chr1, 11, 12, 13 etc. What am I doing wrong?

[pbluescript]

Thanks! I get that this is probably a stupid question, but if I type grep chr1, it pulls chr1, 11, 12, 13 etc. What am I doing wrong?

Use grep -w chr1

[dpryan]

Thanks! I get that this is probably a stupid question, but if I type grep chr1, it pulls chr1, 11, 12, 13 etc. What am I doing wrong?

Nothing, I should have foreseen that! Mea culpa. There's a "word boundary" switch you can give to grep (at least on my computer). So try "grep -w chrX file.vcf > chrX.vcf". That should work better!