Is it possible to use haplotypes instead of SNPs for association testing with phenotypes?

Question

[William]

We need to start thinking about genome analysis as haplotype reconstruction- not variant calling

We need to start thinking about genome analysis as haplotype reconstruction- not variant calling. 5/5

— Deanna Church (@deannachurch) June 20, 2017

Using new technology like linked reads or long reads it is becoming much easier to identify haplotypes in genomes, ie. collections of phased variants.

Can these haplotypes be used for association testing with phenotypes (e.g. qtl mapping, gwas, genomic selection) instead of SNPs? And is there any benefit to using haplotypes? Or will the future of association testing stay with SNPs?

Are there any good articles on this?

[William]

Reference for genomic selection method http://www.genetics.org/content/157/4/1819

Answer 1

[GabrielMontenegro]

If by genomic selection you mean using haplotypes to find signatures of selection, then the answer is yes. Many popular haplotype-based tests are conducted using haplotype data. From the top of my head: iHS, nSL, XP-EHH. There are probably many more.

Doing GWAS with haplotypes is also possible and there are many softwares that can do that. Compared to individual SNP GWAS they have advantages and disadvantages, like any other method. Here's a list that I found:

https://omictools.com/haplotype-association-category

HTH

[William]

Thank you for the haplotype association link. I'll look in to the tools / papers listed there. I added a reference for what I mean with genomic selection as a comment to the orginal question.