Allele Frequency Of Somatic Mutations In Individuals Reveals Signatures Of Cancer-Related Genes

In recent years, comprehensive sequencing efforts have revealed signatures of cancer-related genes, such as driver and passenger genes, oncogenes and tumor suppressor genes(TSGs), which makes a great contribution to cancer research. However, these discoveries are based on hundreds of cancer samples, costing a lot of time and money, ignoring the heterogeneity among different samples, and useless to out-of-rule genes. Here we tried to study the signatures using allele frequency of somatic mutations in individuals. We analyzed exome sequencing data of 50 pairs of primary human colorectal tumors, and identified 5 driver genes(TP53, APC, KRAS, SMAD4, PIK3CA). The order of their allele frequency was almost consistent with their chronological sequence proved by previous studies. Considering that different patients may have different driver genes, we obtained additional 542 potential driver genes from the COSMIC database. Comparative analysis results showed the allele frequency of the driver genes and the potential driver genes in individuals were both significantly higher than that of the passenger genes, indicating that some potential driver genes, which could not be identified by population mutation frequency(MutSig), may also be involved in the progression of CRC. Furthermore, the allele frequency of the oncogenes in individuals was significantly lower than that of the TSGs. Our results suggested that allele frequency of somatic mutations in individuals could be used not only to predict mutations progression, but also to identify driver genes and distinguish oncogenes and TSGs.