difference betwween hail and plink

Question

[kkk]

Hello! I am considering installing a Hail, and I have a question about Hail. I think it is a service similar to plink, but can you tell me what is different and what hail can only do? I would appreciate it if you could be as specific as possible.

Is it possible to do things like MTAG for gwas with both? I am also interested in LinierMixedModel, etc.

Best regards.

[DBScan]

You can find the manual for HAIL here https://hail.is/docs/0.2/index.html and the manual for PLINK here https://www.cog-genomics.org/plink/2.0/.

Answer 1

[Jeremy Leipzig]

PLINK is a mature GWAS library. Hail is a framework for distributed variant queries on Spark. Hail does have some GWAS tools but people who are serious about GWAS generally use SAIGE or Regenie. There is a Spark enabled Regenie implementation called GLOW, but the GLOW codebase has little to no Hail.

[kkk]

Thanks for letting me know. What is it that only hail can do? Other than distributed variant queries. And conversely, what can only plink do, what can only regenie do, what can only SAIGE do?

For example, can any of them do LMM GWAS? multi trait analysis?, rare variant analysis?

All I know is that regenie is good at LMM and hail is good at distributed processing. It seems like plink and hail can be used for a variety of purposes, but regenie is limited?

[Jeremy Leipzig]

Hail https://discuss.hail.is/t/big-picture-issues-considering-switching-to-hail/991/4 SAIGE https://pubmed.ncbi.nlm.nih.gov/30104761/ REGENIE https://pubmed.ncbi.nlm.nih.gov/34017140/

[kkk]

Thank you sooo much! You have givin me great progress!