多组学数据整合分析癌症细胞生存必需基因

Essential genes carry genetic information indispensable for the survival of organisms under certain conditions. Loss of essential genes leads to cell death. Identification of essential genes is an important aspect of understanding basic biological processes, discovering disease-specific effective drug targets, as well as selecting genes for targeting which can potentially synergize with existing treatment regimens. Currently, identification of essential genes is primarily reliant on large-scale genetic knockout/knockdown or transposon mutagenesis methods. While providing functional validation of essential genes uncovered, these methods are expensive and logistically challenging. We posit that computational methods can provide valuable, alternative approaches to facilitate accurate prediction of essential genes. We recently developed an algorithm taking into consideration protein interaction and gene co-expression data for analysis of genes that are essential to tumor development. In this study, we propose to further incorporate protein-DNA interaction (ChIP-seq), chromatin accessibility (ATAC-seq and DNase-seq) and large-scale CRISPR screen data into computational gene essentiality prediction. Candidate essential genes predicted by our algorithm will be validated using CRISPR/Cas9-mediated genetic knockout methods. Out work presents a novel opportunity for systematic identification of novel drug targets and combinations for cancer therapy.

必需基因是指细胞在特定条件下存活所必需的基因，该基因的缺失会导致细胞死亡。寻找细胞生长必需基因对于研究细胞生命基础功能、发现疾病特异靶点和鉴定药物协同作用靶点都至关重要。目前，鉴定必需基因主要依靠基因敲除、基因沉默或者转座子突变技术。这些技术能够精确的识别一些必需基因，但是价格昂贵，并且费时费力。因此通过计算方法来精确预测必需基因有巨大的价值，申请人所在实验室已经能成功通过蛋白相互作用（PPI）和基因共表达信息来鉴定肿瘤中的必需基因。在本课题中，申请将在前期研究的基础上进一步整合ChIP-seq、染色质异质性（ATAC-seq和DNase-seq）和CRISPR高通量筛选数据来优化必需基因的预测模型，并通过CRISPR/Cas9实验技术验证对预测结果进行验证。我们的工作可以用来鉴定癌症中新的药物靶点和寻找新的联合用药方式。

基因表达调控对于理解哺乳动物细胞分裂、分化及疾病状态下细胞内调控机制具有重要研究意义，有助于在分子层面鉴定疾病特异的药物靶点。表观遗传学高通量测序技术的发展，使得人们可以在全基因组尺度具体研究不同生物学系统中，基因表达调控与疾病发生发展的关系。项目申请人前期的工作整合了大量已发表的表观遗传组学数据、癌症基因组学数据，试图探讨基因表达调控对癌症发生发展的调控关系。在本项目的支持下，申请人进一步开展了两方面的工作。一，开发了系统收集、处理、展示表观遗传组学数据的生物信息学软件，并对已发表的表观遗传组学数据进行了收集与处理，构建了CistromeDC数据库。二，基于CistromeDC数据库中收集的表观遗传组学数据，开发了生物信息学方法Lisa，可以基于差异基因表达数据，鉴定调控这些基因的关键转录因子。该方法可以用来鉴定和预测发育、疾病过程中起关键作用的药物靶点。在项目执行期间，负责人发表论文2篇，其中基金标注2篇。