Create ROI files that work generically for neoantigen prediction regardless of exome reagent used

[malachig]

This issue is not specifically related to the immuno.wdl code itself but instead a common input file.

We are routinely processing exome capture data from many different providers, including those where the correct exome reagent definition (exome ROI BED file) is not available. We periodically filter out variants called by another external pipeline and realize that they used a different definition.

One thought would be to create an ROI file that covers every position that could give rise to a neoantigen according to the version of Ensembl being used for the analysis.

[malachig]

I created a version of these that adds any missing coding regions to the IDT exome definition as an example:

gs://griffith-lab-workflow-inputs/human_GRCh38_ens105/capture_reagents/IDT_xGen_Lockdown_Exome_v1.baits.plusEnsemblCDS.interval_list
gs://griffith-lab-workflow-inputs/human_GRCh38_ens105/capture_reagents/IDT_xGen_Lockdown_Exome_v1.targets.plusEnsemblCDS.interval_list