Tan Hua, Zhou Xiaobo
Department of Radiology, Wake Forest School of Medicine, Center for Bioinformatics & Systems Biology, Medical Center Blvd., Winston-Salem, NC, 27157, USA.
Methods Mol Biol. 2018;1711:3-11. doi: 10.1007/978-1-4939-7493-1_1.
Cancer genes may tend to mutate in a co-mutational or mutually exclusive manner in a tumor sample of a specific cancer, which constitute two known combinatorial mutational patterns for a given gene set. Previous studies have established that genes functioning in different signaling pathways can mutate in the same sample, i.e., a tumor from one patient, while genes operating in the same pathway are rarely mutated in the same cancer genome. Therefore, reliable identification of combinatorial mutational patterns of candidate cancer genes has important ramifications in inferring signaling network modules in a particular cancer type. While algorithms for discovering mutated driver pathways based on mutual exclusivity of mutations in cancer genes have been proposed, a systematic pipeline for identifying both co-mutational and mutually exclusive patterns with rational significance estimation is still lacking. Here, we describe a reliable framework with detailed procedures to simultaneously explore both combinatorial mutational patterns from public cross-sectional gene mutation data.
癌症基因在特定癌症的肿瘤样本中可能倾向于以共突变或互斥的方式发生突变，这构成了给定基因集的两种已知组合突变模式。先前的研究表明，在不同信号通路中起作用的基因可在同一样本（即来自一名患者的肿瘤）中发生突变，而在同一通路中起作用的基因在同一癌症基因组中很少同时发生突变。因此，可靠地识别候选癌症基因的组合突变模式对于推断特定癌症类型中的信号网络模块具有重要意义。虽然已经提出了基于癌症基因中突变互斥性来发现突变驱动通路的算法，但仍然缺乏一种用于识别具有合理意义估计的共突变和互斥模式的系统流程。在此，我们描述了一个可靠的框架及详细程序，用于同时从公共横断面基因突变数据中探索这两种组合突变模式。
