1
$\begingroup$

I have a list of SNPs (index SNPs). As my tagging SNPs, I want to use SNP data from here. The data has the following column headers: Chromosome, Position, MarkerName, Effect_allele, Non_Effect_allele, Beta, SE, Pvalue.

My ultimate aim is to identify SNPs in the tagging set that are in LD with each of index SNPs. I need to calculate r2 between index SNPs and tagging SNPs within 1MB window and get a table of index vs tagging SNPs (r2>0.8).

How can I use PLINK to carry out this work? I'm new to PLINK and LD clumping

Improve this question
$\endgroup$

1 Answer 1

Reset to default
1
$\begingroup$

You can't generate LD statistics for summary data from SNPs, as the LD calculation relies on the comparison of variants at the genotype level. Plink expects data to be in a genotype matrix with either each marker for each line (probably better in your case), or each individual for each line. Plink v1.9, in particular, will accept VCF-formatted input files, when populated with 'GT' information that describes the variants for each individual.

If you're okay with using the 1000 genome population for calculating LD, plink v1.9-compatible datasets should be accessible from here.

If not, you'll need to discuss your project with IGAP to see if you can have access to genotype-level data for LD testing. That's unlikely, because there are personal identification issues associated with non-summary data.

There's comprehensive documentation about the LD command(s) on the Plink 1.9 website here.

Improve this answer
$\endgroup$
2
  • $\begingroup$ Thanks for your explanation @gringer. I will now use a plink v1.9-compatible dataset from 1000 genome and then use plink --noweb --bfile 1000genomefile --clump SNPfile --clump-field P --clump-p1 1.0001 --clump-p2 1.0001 --clump-r2 0.8 --clump-kb 1000 --out resultfile. But I'm unsure if line-by-line listing of SNPs on SNPfile will work and what the correct format should be. Any suggestions? $\endgroup$
    – user27976
    Nov 22, 2017 at 12:48
  • $\begingroup$ EDIT: --clump-p1 0.0001 --clump-p2 0.0001 $\endgroup$
    – user27976
    Nov 22, 2017 at 14:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.