Is your feature request related to a problem? Please describe.
We need a way to take VCFs and using ONLY GUIs get it into CRUX.
Describe the solution you'd like
Ideally, just dropping in lots of single sample VCFs or 1 multi sample VCF into CRUX would just work. Since we absolutely need variants annotated to make a valid MAF - the input will have to be VEP annotated VCFs.
Because VCF2MAF conversion requires certain decisions to be made (e.g. how to figure out Tumour Sample Barcodes when using single sample VCFs? Should any filtering be applied (e.g. based on population AF), etc. the configuration of a VCF2maf might be complex enough that its simpler to just implement in interchangeGUI
Additional context
If we tell a user to annotated using graphical VEP interface - there are no guarantees VEP version and thus gene-names / impact classifications will be concordant with the datasets precompiled in CRUX. We should warn about this - so we don't mislead people into doing 2-cohort comparisons of custom VS inbuilt datasets with different VEP versions (gene name changes would look like cohort specific mutation patterns). Commandline VCF2maf with a specified VEP version can guarantee equivalency
Is your feature request related to a problem? Please describe.
We need a way to take VCFs and using ONLY GUIs get it into CRUX.
Describe the solution you'd like Ideally, just dropping in lots of single sample VCFs or 1 multi sample VCF into CRUX would just work. Since we absolutely need variants annotated to make a valid MAF - the input will have to be VEP annotated VCFs.
Because VCF2MAF conversion requires certain decisions to be made (e.g. how to figure out Tumour Sample Barcodes when using single sample VCFs? Should any filtering be applied (e.g. based on population AF), etc. the configuration of a VCF2maf might be complex enough that its simpler to just implement in interchangeGUI
Additional context If we tell a user to annotated using graphical VEP interface - there are no guarantees VEP version and thus gene-names / impact classifications will be concordant with the datasets precompiled in CRUX. We should warn about this - so we don't mislead people into doing 2-cohort comparisons of custom VS inbuilt datasets with different VEP versions (gene name changes would look like cohort specific mutation patterns). Commandline VCF2maf with a specified VEP version can guarantee equivalency