Description
Many studies have identified cancer subtypes based on the cancer driver genes, or
the proportion of mutational processes in cancer genomes, however, none of these cancer subtyping
methods consider these features together to identify cancer subtypes. Accurate classification of
cancer individuals with similar mutational profiles may help clinicians to identify individuals who
could receive the same types of treatment. Here, we develop a new statistical pipeline and use a
novel concept, “gene-motif”, to identify five pancreatic cancer subtypes, in which for most of them,
targeted treatment options are currently available. More importantly, for the first time we provide
a system-wide analysis of the enrichment of de novo mutations in a specific motif context of the
driver genes in pancreatic cancer. By knowing the genes and motif associated with the mutations, a
personalized treatment can be developed that considers the specific nucleotide sequence context of
mutations within responsible genes.
