Interrogating Epigenetic Changes in Cancer Genomes

描述（由申请人提供）：俄亥俄州州立大学的研究者提交了一份中心资助申请，以回应RFA宣布的综合癌症生物学项目（ICBP）。 此外，我们在邻近的印第安纳州大学的同事将加入这一奋进。 拟议的研究将解决一个独特的问题，即，表观遗传改变，现在被认为与癌症中的基因突变一样重要。 这种现象可以被定义为一种可遗传的变化，调节染色质组织和基因表达，而不改变核苷酸序列。 ICBP汇集了实验和计算生物学家，我们的总体目标是1）增加我们对肿瘤中复杂表观遗传相互作用的理解，2）使用高端信息改善人类女性癌症的预后，干预和治疗。 实验生物学家将使用新的微阵列平台来询问癌细胞系和肿瘤上皮及周围基质中的DNA甲基化、组蛋白修饰、杂合性丢失和转录因子结合。 计算生物学家将使用这些实验数据进行模型构建和改进。 经验贝叶斯模型将用于预测如何招募阻遏物，组蛋白脱乙酰酶和DMA甲基转移酶，以建立表观遗传基因沉默（项目1）。 将开发系统发生聚类算法，以概括癌症间质中与肿瘤进展相关的遗传和表观遗传途径（项目2）。 LASSO逻辑回归和神经网络方法将用于对协同DNA-蛋白质相互作用和癌细胞中染色质景观的变化进行建模（项目3）。 模式识别和监督学习技术将用于选择耐药癌细胞中含有甲基化倾向序列特征的基因（项目4）。 这些数学模型将产生第一级或第二级假设，用于实验检验。模型改进和实验验证的迭代过程将继续下去，直到得到准确预测询问癌症基因组中特定表观遗传改变的模型。 所有的实验数据和建模工具都将存储在一个集中的数据库中，并将用于培训未来的系统癌症生物学家。 这些综合研究的进展将由顾问和行政领导人进行评估，以确保拟议的ICBP取得成功。

DESCRIPTION (provided by applicant): The investigators of the Ohio State University submit a center grant application in response to the RFA announcement of the Integrated Cancer Biology Program (ICBP). In addition, our colleagues at the neighboring Indiana University will join this endeavor. The proposed research will tackle a unique problem, i.e., epigenetic alteration, now being considered as important as genetic mutations in cancer. This phenomenon can be defined as a heritable change that modulates chromatin organization and gene expression without altering nucleotide sequences. Bringing together experimental and computational biologists, our overall goals of this ICBP are 1) to increase our understanding of complex epigenetic interactions in neoplasms and 2) to use high-end information for improved prognosis, intervention, and treatment of human female cancers. Experimental biologists will use novel microarray platforms to interrogate DNA methylation, histone modifications, loss of heterozygosity, and transcription factor binding in cancer cell lines and neoplastic epithelium and the surrounding stroma. Computational biologists will use these experimental data for model building and refinement. Empirical Bayesian models will be used to predict how repressors, histone deacetylases, and DMA methyltransferases are recruited to establish epigenetic gene silencing (Project 1). Phylogenetic clustering algorithms will be developed to recapitulate genetic and epigenetic pathways in cancer stroma as they relate to tumor progression (Project 2). LASSO logistic regression and neural network approaches will be used to model the synergistic DNA-protein interactions and the resulting change of chromatin landscape in cancer cells (Project 3). Pattern recognition and supervised learning techniques will be used to select genes that contain the characteristics of methylation-prone sequences in drug-resistant cancer cells (Project 4). These mathematical models will generate the first- or second-level hypotheses for experimental testing. The iterative process of model refinement and experimental verification will continue until models are derived that accurately predict specific epigenetic alterations in the interrogating cancer genome. All experimental data and modeling tools will be deposited in a centralized database and will be used for training future systems cancer biologists. The progress of these integrated studies will be evaluated by advisors and administrative leaders to ensure the success of the proposed ICBP.