基于多调控因子的细胞/组织特异性基因调控机理研究

Precision Medicine, emphasizing the individual characteristics of each patient, requires the deep understand on the disease associated gene regulatory mechanism. However, this task is complicated since multiple factors are involved in the gene regulation, among which the interrelations were reported. Benefited from the high-throughput sequencing, the techniques for chromatin structure and epigenomic detection provide new insight into the gene regulatory researches. Meanwhile the development in the deep-learning makes the possible for integrative analysis on diversity of biological data. Here, we would focus on the cell/tissue specificity related gene regulatory. On basis of the deep-learning and complex network algorithms, we plan to make investigation into the correlation between epigenetic modifications, chromatin structures and transcript factors. And then design algorithms for assessing gene expression level by taking all the factors into account. We then ask the question how these factors organized in the regulatory process. By employing the complex network algorithm, we could explore the combinational patterns between these factors, called as motif, for their function. Finally, these function related motifs would be applied into the cancer study, such as biomarker and drug target screening. The strategy proposed in this project for studying gene regulation by multiple factors fusing could provide new insight into the disease associated molecular mechanism studies.

精准医疗被认为是划时代的革命，其强调的个体差异要求对疾病相关基因的调控机理有深入了解。但基因调控因子的多样化以及各因子之间的关联等特点对研究工作者提出了巨大挑战。高通量测序技术尤其是以此为基础提出的染色质结构和表观基因组检测技术为基因调控的研究提供了新的视角。另一方面，深度学习算法的发展为生物大数据的融合分析提供了方法学上的支撑。本项目拟围绕细胞/组织特异性的基因调控机理，借助深度学习和复杂网络算法，开展表观修饰、染色质结构和转录因子之间的关联研究；进而从多个层次探索基因表达机理；以此为基础构建基因调控网络，挖掘功能模体，并尝试应用于癌症的生物标记物筛选和药物靶标筛选。该项目中关于对基因调控的多层次融合分析的策略能够为疾病机理研究提供新的思路。

精准医疗被认为是划时代的革命，其强调的个体差异要求对疾病相关基因的调控机理有深入了解。但基因调控因子的多样化以及各因子之间的关联等特点对研究工作者提出了巨大挑战。高通量测序技术尤其是以此为基础提出的染色质结构和表观基因组检测技术为基因调控的研究提供了新的视角。另一方面，深度学习算法的发展为生物大数据的融合分析提供了方法学上的支撑。本项目拟围绕细胞/组织特异性的基因调控机理，借助深度学习和复杂网络算法，开展表观修饰、染色质结构和转录因子之间的关联研究；进而从多个层次探索基因表达机理；以此为基础构建基因调控网络，挖掘功能模体，并尝试应用于癌症的生物标记物筛选和药物靶标筛选。该项目中关于对基因调控的多层次融合分析的策略能够为疾病机理研究提供新的思路。