Sites dataset prep pipeline (subset data)

[ignatiusm]

Description

For Autism CRC data, we need a pipeline to create the required input data. This involves converting gnomad_qc v2 data prep steps into a form that they can be applied using a dataproc cluster, similar to other gnomAD data steps.

Definition of done

• ensure Matthew has access to Garvan forks of gnomad_qc, gnomad_methods and gnomad-browser repos
• have meeting with Matthew re Nature paper and methods descriptions
• converted the gnomad_qc v2 data prep steps into functions
• created a gnomad-browser pipeline that can run v2 data prep steps on subsetted data
• Have matthew review the results

Alternative Approach

I'm trying to work backwards (to see which data prep steps we can skip) by trying to load data into the backend, and then progressively including more of the data loading steps in reverse order:

• CAIDS (ClinGen Canonical Allele IDs) prep step
• MNVS (Multinucleotide variants)