We noticed that some positions missed by SNVMix1 at 40× were in what we believe are allele-specific copy number amplifications. Future work will involve incorporating copy number data directly into the model to consider such situations where the resultant allelic bias is expected to mask variants present in the unamplified allele. These results will be presented in a forthcoming manuscript. In a similar vein, due to regulatory mutations or epigenetic changes, transcriptomes can show preferential expression of one allele and thus our model will be insensitive to instances of extensively skewed allelic expression. Further extensions of the model to consider these factors will be explored.
