Hail SVCR local alleles

[hammer]

Hail’s multi-sample storage strategy gets a preprint: https://www.biorxiv.org/content/10.1101/2024.01.09.574205v1

I’m not sure this changes much for our paper though we should probably point to it in the discussion and also look into mapping it onto Zarr someday!

[jeromekelleher]

Interesting - agree we should discuss, and it doesn't really change our narrative

[jeromekelleher]

Noting here the the "local alleles" idea from SVCR is a elegant and powerful idea that we should adopt. I'm currently working on VCF conversion, and the PL fields are a serious challenge to the current model. I've got a 1000 genomes example here:

 "shape": [  
        2958525,
        3202,
        28
    ],      

which is pretty impractical.

[jeromekelleher]

Local alleles has also been implemented in bcftools apparently, as mentioned in discuss of the VCF perspective: https://academic.oup.com/gigascience/article/10/2/giab008/6137722