Computational Methods for Identifying Non-coding Cancer Drivers

从肿瘤全基因组序列(WGS)获得的大多数变异发生在非 基因组的编码区。尽管蛋白质编码区的变体已经收到了 大多数人注意到，许多研究现在都注意到了非 癌症中的编码变体。致癌功能非编码变异体的鉴定 增长仍然是全基因组测序应用的一个挑战和瓶颈 诊所。癌症驱动因素通常是由癌症的高频率来识别的 突变会在患者中发生。然而，在非传染性疾病中，突变率具有高度的异质性。 编码区和许多非驱动元件的突变频率高于 其他的，如黑色素瘤中转录因子结合的区域或区域 结肠癌细胞分裂后期复制。在本提案中，我们将使用高- 吞吐量池CRISPR筛选和预测非 对癌症驱动因素进行编码。我们将定量衡量数以千计的非 使用我们创新的高通量CRISPR屏幕编码突变，直接连接 在非编码基因组的天然背景下的修改(即不是报告分析) 与癌症相关的表型(细胞生长)。筛选的结果将用作 用于开发计算癌症驱动程序预测的NC_DRIVER的训练数据 工具。NC_DRIVER将高功能影响的信号与重复性 跨多个肿瘤样本的变异以识别以下非编码突变 癌症中的阳性选择。我们将确定推动者、增强者和CTCF中的驱动因素 绝缘体。CTCF绝缘子是变异最多但研究最少的调节元件 在癌症基因组中。使用这种实验和计算相结合的方法， 我们将确定高信心的候选司机。最后，我们将执行函数 结直肠癌和前列腺癌中优先非编码驱动因素的评估。我们会 在患者来源的细胞培养中使用CRISPR/Cas9基因组编辑来测试20个高排名 候选驱动程序启动子/增强子/绝缘子突变。总体而言，这项建议 解决了识别非编码基因组中的驱动因素以及长期- TERM使每个患者的基因组测序受益最大。

Most variants obtained from tumor whole-genome sequences (WGS) occur in non- coding regions of the genome. Although variants in protein-coding regions have received the majority of attention, numerous studies have now noted the importance of non- coding variants in cancer. Identification of functional non-coding variants that drive tumor growth remains a challenge and a bottleneck for the use of whole-genome sequencing in the clinic. Cancer drivers are generally identified by the high frequency at which their mutations occur across patients. However, mutation rate is highly heterogeneous in non- coding regions and many non-driver elements show higher mutation frequency than others, such as regions bound by transcription factors in melanoma or regions replicating late during cell division in colon cancer. In this proposal, we will use high- throughput pooled CRISPR screen and novel computational methods to predict non- coding cancer drivers. We will quantitatively measure the impact of thousands of non- coding mutations using our innovative high-throughput CRISPR screen that directly ties modifications in the native context of the non-coding genome (i.e. not a reporter assay) to a cancer relevant phenotype (cell growth). The results of the screen will be used as training data for the development of NC_Driver, a computational cancer driver prediction tool. NC_Driver will integrate the signals of high functional impact with the recurrence of variants across multiple tumor samples to identify the non-coding mutations under positive selection in cancer. We will identify drivers in promoters, enhancers and CTCF insulators. CTCF insulators are the most mutated yet least studied regulatory elements in the cancer genome. Using this integrative experimental and computational approach, we will identify high-confidence candidate drivers. Finally, we will perform functional evaluation of prioritized non-coding drivers in colorectal and prostate cancers. We will use CRISPR/Cas9 genome editing in patient-derived cell cultures to test 20 high-ranking candidate driver promoter/enhancer/insulator mutations. Overall, this proposal addresses the critical need to identify drivers in the non-coding genome and over long- term enable the maximal benefit of genome sequencing for each patient.