Haploview LD plot unespected results

Question

[chiuni2308]

Hi, I'm analyzing linkage disequilibrium (LD) between 3 SNPs in the VDR gene using Haploview. The SNP positions are:

TAQI: 47844974 APAI: 47845054 BSMI: 47846052

Based on physical distance, APAI is much closer to TAQI (80 bp apart) than to BSMI (998 bp apart). I would therefore expect APAI to show higher LD with TAQI. However, Haploview shows that APAI and BSMI have higher LD than APAI and TAQI.

My question is: what biological or statistical factors could explain this result? Is it possible for two SNPs further apart physically to have higher LD due to recombination rates or haplotype structure?

Any insight or references to similar cases would be greatly appreciated.

Here is the LD plot:

[Kevin Blighe]

Nobody can really answer this properly. You are looking at 3 genetic variants, whereas, in this type of analysis, we are usually looking at dozens or more - see: How do I compute ld blocks from the hapmap ld_data?

Answer 1

[Prash]

I think there could be several reasons, IMHO

This long range LD problem could be because of epistasis and population admixture

I also work(ed) on GC and VDR and this is expected as long as you compared LDs across population admixture

May I suggest you use LDkit as an alternative if this is not inviting a bias

Hope this helps, Prash

[Kevin Blighe]

I agree with you.