VCF - Genome Nexus annotation script

[n1zea144]

Background -

AACR Project GENIE collects VCFs of various flavors from consortium members. Currently they use vcf2maf to annotate these files. We desire to move them to the Genome Nexus platform. There we would like to create a script for annotating a set of VCF files using our GN annotation pipeline. Rough outline:

Point tool to folder which contains VCFs of various shapes/size.
Run vcf2vcf over files to normalize them.
Run normalized vcfs through @ao508 vcf2maf converter.
Annotate mafs using genome nexus annotation client.
Combine mafs into single file for import.
Note, for 2:
vcf2vcf is required because there is a need to normalize allele depths which is handled by FixAllelDepths. Ensure the following option is set (should be set by default)

--retain-format GT,AD,DP

These three fields contain the values needed to extract t_depth, t_ref_count, t_alt_count, n_depth, n_ref_count, n_alt_count.

https://github.com/mskcc/vcf2maf

[sheridancbio]

one comment .. it was observed in the running of vcf2vcf that explicitly setting --retain-format GT,AD,DP cause the "loss" of the genotype column values. When this argument was unused, the output was better.

[sheridancbio]

This card was originally for the implementation of the entire scripting / processing pipeline. In re-planning this card, we are re-using this card now to represent only the basic updates to the existing vcf tools (vcf2vcf and vcf2maf) so that VEP is no longer run, and the tools run smoothly without errors or perl warnings (such as the use of uninitialized variables during output generation)

[sheridancbio]

This card was completed with this PR : mskcc/vcf2maf#246