Hyper-mutational processes provide a head-start for non-optimal cancer driver mutations explaining atypical KRAS variants

In hypermutant cancers, both the mutation rate per cell division and the genomic locations of new mutations are vastly altered. Such cancers often display atypical driver gene variants, which may provide sub-optimal oncogenic advantages. Why such weaker cancer drivers prevail in these tumours is unclear. Here, we examine the null hypothesis that aberrant mutational processes alone can account for the detection of weaker, rather than canonical, driver mutations. Using a mathematical modelling analysis, we find that simply increasing the mutation rate per-division can lead to the detection of weaker drivers, an effect which is further compounded by the biassing of mutations towards weaker driver sequence contexts. Focusing on POLE-mutant (DNA polymerase epsilon proofreading-deficient) colorectal cancers and KRAS driver mutations, we quantify the mutational bias towards non-canonical KRAS drivers in both POLE-mutant and non-hypermutant colorectal cancers. The observed bias, coupled with the increased mutation rate of POLE-mutant cancers, is sufficient to explain the enrichment of atypical KRAS drivers in POLE-mutant cancers with plausible selection parameters. Thus, within our model, differential selection between these cancer types need not be invoked to understand the observed variation in driver mutations.