How to perform Linkage Disequilibrium analysis for significantly associated K-mers?

Question

[anikcropscience]

Hello, I am performing a K-mer based GWAS on my dataset. I have obtained some k-mers that are significantly associated with the trait of interest. Now, I want to perform a linkage disequilibrium (LD) analysis among those significant K-mers and create a heatmap as it is done with SNPs. But the problem is that the positioning of the k-mers is not know as it is a reference-free method. I have a table with presence/absence pattern of the associated k-mers in my samples.

Does someone have any idea how can I measure the LD among the significant k-mers?

I have the Plink binary files (*.bim, *.bed and *.fam files) containing the K-mer data that I used for GWAS.

Please let me know if you have any ideas.

Thank you.

Answer 1

[biostarter]

what tool did you use for your gwas and how was the presence/absence kmer pattern generated?

[anikcropscience]

I used the K-mer GWAS tool to do the GWAS [https://www.nature.com/articles/s41588-020-0612-7]

If one gets any significant association between a K-mer and the phenotype, then a presence/absence pattern of that K-mer among the samples can be obtained. The presence/absence pattern is in binary form. I saw in the paper that they used a formula for getting the LD value under the section "Calculation of Linkage Disequilibrium". I asked the author about the input files. They said they only used that presence/absence pattern table and applied that formula in R to obtain the LD value. But I am not sure how to use that formula for estimating LD in R. Do you have any idea?

Answer 2

[srividhyasainath2]

Hi Anik,

Did you manage to calculate LD between significant kmers? if yes, which tool did you eventually use and what were the input files for the same.

Thanks