Suggestion to add imputation by reference allele

[mfasold]

I think a very useful addition would be an option that allows users to calculate PRS with imputation based on the reference allele. As I understand, for all variants not contained in the input VCF, the frequency of the effect allele in the selected population is used. When you receive an VCF from a WGS provider, it will most likely contain only the non-reference alleles, whereas for the overwhelming rest it can be assumed that they are reference alleles. In the future, also a gVCF file could be used to differentiate between unknown and reference call.

[miller34]

Thanks for the suggestion. Yes. That would be a great addition to include gVCF support. Imputation is currently calculated based on the variant alleles and does not consider the reference alleles.