While recently emergent driver mutation data sets are available for developing computational methods to predict cancer mutation effects, benchmark sets focusing on passenger mutations are largely missing. Here, we developed a comprehensive literature-based database of Cancer Passenger Mutations (dbCPM), which contains 941 experimentally supported and 978 putative passenger mutations derived from a manual curation of the literature. Using the missense mutation data, the largest group in the dbCPM, we explored patterns of missense passenger mutations by comparing them with the missense driver mutations and assessed the performance of four cancer-focused mutation effect predictors. We found that the missense passenger mutations showed significant differences with drivers at multiple levels, and several appeared in both the passenger and driver categories, showing pleiotropic functions depending on the tumor context. Although all the predictors displayed good true positive rates, their true negative rates were relatively low due to the lack of negative training samples with experimental evidence, which suggests that a suitable negative data set for developing a more robust methodology is needed. We hope that the dbCPM will be a benchmark data set for improving and evaluating prediction algorithms and serve as a valuable resource for the cancer research community. dbCPM is freely available online at http://bioinfo.ahu.edu.cn:8080/dbCPM.
