Quantitative Modeling for Chromatin Regulation of Gene Expression in Cancer

项目摘要 染色质在转录调控中起重要作用。染色质相关基因经常发生突变 在癌症中。深入研究染色质在基因调控中的作用，对于理解染色体的功能， 肿瘤发生和肿瘤进展的分子机制。作为一名经验丰富的计算生物学家， 在ChIP-seq生物信息学和表观遗传学方面的专业知识，我的研究重点是开发计算 染色质调控的高通量基因组数据分析和计算建模方法 的基因表达。在癌症生物学方面有了更多的独立研究训练，我将发展我的研究 计算机癌症表观遗传学计划，并发展独立的学术生涯。 最近的研究已经证明了针对细胞中活性开放区域的染色质调节剂的可行性。 基因组作为癌症治疗的新疗法。然而，染色质的背景特异性底物 调节子和染色质如何调节基因表达的潜在机制在很大程度上还不清楚。与 基于下一代测序的高通量基因组技术（包括ChIP-seq）的出现， DNase-seq和ATAC-seq，大量的基因组分析数据变得可用， 用综合计算方法系统地破译基因调控机制。 本项目的目标是开发新的定量和计算方法，用于研究 表观遗传基因调控和染色质调节因子在癌症中的功能。具体而言，我们建议 开发综合计算方法，利用丰富的公共ChIP-seq、DNase-seq， 以及用于预测功能性调节元件和TF的ATAC-seq数据。首先（目标1），我们将开发 一种方法，预测功能增强子元件和相关的TF给定任何基因集使用公共 跨多种细胞类型组蛋白标记ChIP-seq数据。第二（目标2），我们将开发一个定量模型， 从配对末端DNase-seq或ATAC-seq鉴定核苷酸分辨染色质可及性动力学 对数据中的固有偏差进行校正。最后（目标3），我们将整合公共可用的DNase- seq、ATAC-seq和ChIP-seq数据，并系统地表征 染色质调节因子的功能，重点是EZH 2在一些癌症系统中的作用，包括去势抵抗性 前列腺癌（CRPC）细胞和恶性外周神经鞘瘤（MPNST）。这些计算 方法补充了现有的生物信息学方法，并将在以下研究中得到广泛的应用： 癌症表观遗传学和基因调控。这项研究将填补致癌基因之间的知识空白。 驱动程序和下游基因表达程序，并可以为开发 癌症精准医学的新型靶向疗法。

PROJECT SUMMARY Chromatin plays an essential role in transcriptional regulation. Chromatin-related genes are frequently mutated in cancers. Dissecting the functions of chromatin in gene regulation is important for understanding the molecular mechanisms of oncogenesis and tumor progression. As an experienced computational biologist with expertise on ChIP-seq bioinformatics and epigenetics, my research has focused on developing computational methodologies for high-throughput genomic data analysis and computational modeling on chromatin regulation of gene expression. With more independent research training in cancer biology, I will develop my research program on computational cancer epigenetics and develop an independent academic career. Recent studies have demonstrated the feasibility of targeting chromatin regulators for active open regions in the genome as novel therapeutics for cancer treatment. However, the context-specific substrates of chromatin regulators and the mechanisms underlying how chromatin regulates gene expression are largely unclear.. With the advent of next-generation sequencing based high-throughput genomic techniques including ChIP-seq, DNase-seq, and ATAC-seq, a large amount of for genomic profiling data became available, making it possible to systematically decipher the gene regulatory mechanisms with an integrative computational approach. The objective of this project is to develop novel quantitative and computational methodologies for studying epigenetic gene regulation and the functions of chromatin regulators in cancer. Specifically, we propose to develop integrative computational methods that leverage the abundant public ChIP-seq, DNase-seq, and ATAC-seq data for predicting functional regulatory elements and TFs. First (Aim 1), we will develop a method that predicts the functional enhancer elements and associated TFs given any gene set using public histone mark ChIP-seq data across multiple cell types. Second (Aim 2), we will develop a quantitative model to identify the nucleotide-resolution chromatin accessibility dynamics from paired-end DNase-seq or ATAC-seq data with correction of intrinsic biases in the data. Finally (Aim 3), we will integrate publicly available DNase- seq, ATAC-seq, and ChIP-seq data in a comprehensive database and systematically characterize the functions of chromatin regulators with a focus on EZH2 in a few cancer systems, including castration-resistant prostate cancer (CRPC) cells, and malignant peripheral nerve sheath tumors (MPNSTs). These computational methods complement existing bioinformatics methodologies and will have broad applications in the study of cancer epigenetics and gene regulation. The proposed research will fill the knowledge gap between oncogenic drivers and downstream gene expression program, and could provide mechanistic support for development of novel targeted therapeutics for cancer precision medicine.