QC section

[hammer]

Hey @benjeffery I am digging through some early sgkit discussions and wanted to pull out some links that might be useful for your work on the QC section. This may all be old news to you so feel free to ignore!

• Poor performance for QC filtering on medium sized genotype data: early motivation for moving away from Hail. Note the pointer to https://github.com/broadinstitute/gnomad_qc which has some representative QC tasks we might consider.
• https://github.com/related-sciences/gwas-analysis/issues/26 has some early thinking and benchmarks related to our https://github.com/pystatgen/sgkit/issues/31. This method could be a good illustrative operation as mentioned by Jerome in the call yesterday, so I figured lots of detail could be helpful.
• Another potential illustrative operation could be PC-Relate. @ravwojdyla wrote up a nice summary of his work at https://github.com/pystatgen/sgkit/discussions/706 that you may not have seen. PCA was an early user story for us (https://github.com/pystatgen/sgkit/issues/95), and our implementation had some interesting wrinkles (https://github.com/pystatgen/sgkit/pull/262).

[benjeffery]

Thanks @hammer, good stuff.