In past reporting periods, we procured a series of primary endometrial carcinosarcomas and paired non-tumor tissues through the Cooperative Human Tissue Network, which is supported by the National Cancer Institute, and exome sequenced a subset of these tumor-normal pairs to whole exome sequencing at the NIH Intramural Sequencing Center. We analyzed these data to identify from somatic variants (present exclusively in the tumor DNA), which we then orthogonally validated using Sanger sequencing. We prioritized a subset of nonsynonymously, somatically mutated genes for further analysis in a two-part mutation prevalence screen. In the current reporting period we: 1. Defined the mutation frequency, and mutation spectrum of the genes in the first phase of the mutation prevalence screen and identified major gaps in sequence coverage for re-analysis. 2. Used statistical methods to determine whether any of the genes in the mutation prevalence screen are so-called significantly-mutated genes, i.e. whether they are mutated at a statistically significantly higher rate than the background mutation rate for endometrial carcinosarcomas. We hypothesize that significantly-mutated genes are likely to be causal cancer genes that contribute to the development of endometrial carcinosarcomas. In the incoming reporting period we will: 1. Determine the so-called coverage of the mutation prevalence screen and sequence over any remaining gaps. 2. Extend the study to the second phase of the mutation prevalence screen in which candidate cancer genes will be sequenced in an additional cohort of endometrial carcinosarcomas.
