Detect the Runs of Homozygosity with the WGS data

Question

[zhang248]

Hi,

I'm using whole genome sequencing data to detect the ROH. I want to know which tool should I use, Plink or bcftools.

Thanks!

Answer 1

[Kevin Blighe]

If you already have VCF or BCF, then BCFtools may be better, and it even has a publication specifically for its ROH capabilities: BCFtools/RoH: a hidden Markov model approach for detecting autozygosity from next-generation sequencing data

If you want to use PLINK, then you will have to first convert your data to PLINK format. That said, PLINK has a more broad set of functions for filtering and QC of genetics data; so, would be a good choice.

My ultimate recommendation for you is to try both.

Kevin

PS - take a look here at these, too:

• ROH ( Runs of Homozygosity Analysis)
• Detection of Runs of Homozygosity from Whole Exome Sequencing Data: State of the Art and Perspectives for Clinical, Population and Epidemiological Studies

[zhang248]

Thanks Kevin, I have tried the Plink, I have tried Plink, but it seems that bcftools is better for detect the ROHs of whole genome sequence data according to the Bcftools/ROH paper and a review https://www.nature.com/articles/nrg.2017.109.pdf. How do you think about this?

[Kevin Blighe]

Hello, also from my experience, BCFtools gave results that appeared to suit my hypothesis (and the hypothesis of my colleagues) on the data that we were analysing.

ROH analysis is not a primary aim of PLINK, of course, and PLINK has its own niche in association studies and family analyses. Undoubtedly, PLINK is one of the greatest programs of all time.

^ the diplomatic answer

[zhang248]

Thank you, I'll try the BCFtools.