validity check code of VEP annotations in protein-coding genes

[KoalaQin]

1. Downloaded the ensembl 101 & 105 gene intervals and parsed them to Hail format, and stored as reference data;
2. Code to count the total number of variants and the number of variants annotated as "protein-coding" biotype in each interval of all the protein-coding genes, to confirm if a gene is not covered by gnomAD release data, or there might be an annotation issue.
3. Confirmed no issue in VEP105 annotation for v4 exomes & genomes, but 4 genes were mistakenly annotated by VEP 101 in v3 genomes to having no protein-coding variant in a protein-coding;
4. ~50-70 genes were not covered by the release data, maybe due to low coverage in these gene regions?

[jkgoodrich]

Some initial thoughts for modifications

[KoalaQin]

put the count functions inside: gnomad_qc/v4/annotations/generate_variant_qc_annotations.py

[jkgoodrich]

Some suggestions to simplify the code and maybe make it just a little faster.

[KoalaQin]

@jkgoodrich I made changes to your suggestions, back to your 2nd round, thx!

[jkgoodrich]

Just a few more things

[jkgoodrich]

Just one more small thing then it LGTM

[jkgoodrich]

LGTM

[KoalaQin]

resources imported with:
hailctl dataproc submit qh1 /Users/heqin/PycharmProjects/gnomad_methods/gnomad/resources/import_resources.py grch38.ensembl_interval.101

hailctl dataproc submit qh1 /Users/heqin/PycharmProjects/gnomad_methods/gnomad/resources/import_resources.py grch38.ensembl_interval.105