Sites table generation

[michael-harper]

To generate our final callset we will need to generate a list of high quality sites specific to our datasets (both exomes and genomes) that can be used to infer relatedness and ancestry. gnomAD v4 has performed this and while we cannot use their final sites table as a resource (due to the nature of our cohorts and ultimate goal of finding novel sites in certain populations) we can follow their process of developing a high quality sites table.

A couple of points:

• I'm unsure if this should exist in the prod-pipes repo or the references repo mainly because I'm unsure if it's possible to run the script from within that repo
• IntervalQC should be performed prior to this step to ensure high quality regions of exome data are used for sites analysis
• The script used to generate the sites table uses the function get_qc_mt which provides all of the parameters and supplementary Hail functions to implement the filters.
  • @KatalinaBobowik has done this previously based on gnomAD v3 parameters though only using genomes from HGDP and 1KG. The current approach will be using genomes and exomes from our own cohorts.

[michael-harper]

@katiedelange