Getting a VCF file from plink .bed, .fam and .bim files
0
0
Entering edit mode
8 weeks ago
hi.there • 0

Hello there,

I need to produce a VCF file from a plink .bed, .fam and .bim file for CrossMap. Would anybody know how to do this?

I found this simple Biostars post: Convert plink files to VCF: Reference allele file ...but am unsure if the format of the above three plink files is correct to run the exact same process. It seems I may need intermediary steps of plink commands of getting a .ped and .map file from the above three. Thank you.

0
Entering edit mode

Did you try following that post? Why makes you think "am unsure if the format ... is correct ..."?

0
Entering edit mode

Thanks Ram. It says the requisite input file is "your_ped_map_input". I wasn't sure if that is the common name of the orig inal 3 plink files. Is it?

0
Entering edit mode

Did you read the plink manual and see what --file expects? Please don't make me do all the work for you. Dig deeper and keep trying + making mistakes, that's how you'll get to your result.

0
Entering edit mode

Thanks Ram. Yes. I have read all of this documentation before. --file and --bfile. I have actually tried doing the complete liftover before as well with chain files, .ped and .map generation, UCSC BED file generation, awk commands etc. Unfortunately, the last attempt cobbled together resulted in the experimental difference between liftover and no-liftover being the same (which should be impossible as the biomarker we were looking for is not in the old assembly). That's why I'm back at square one here because something somewhere in the pipeline went wrong last time lifting over.

I have: plink2 --bfile [prefix] --export vcf --out <prefix>

I would like to confirm that this is correct.

0
Entering edit mode

The command looks fine and should theoretically work. How is it different from the one you ran last time?

0
Entering edit mode

Thanks Ram. I appreciate you taking the time. I didn't generate a .vcf last time. In the last attempt, I tried updating based on chr position with an all-snp.txt file from UCSC and then lifting over the remaining rsids. It didn't work but I gather it's because the pipeline came together from 6 different people and I was a novice and probably made a mistake somewhere in combining all of their recommendations .