My friend has some WGS data and is trying to identify the frequency of clinically actionable pharmacogenomics variants in specific genes, e.g. CYP2C19*9. Does she need to do phasing before working out the diplotypes? What is the best way of calling the clinically actionable PGx variants from WGS? Can you help my friend?
Phasing is definitely important when trying to annotate star alleles that don’t have a unique allele defining variants. It may be best to use one of the pharmacogenomic tools designed for WGS that supports your genes of interest. Many of these tools were initially designed for CYP2D6, but now support other pharmacogenes. PharmCAT is another tool for assigning clinically-relevant haplotypes, although I have not used it personally.