Genetic variation and regulatory networks: Mechanisms and complexity

建议的研究重点是了解序列变异对基因的影响 分子网络的功能。我们将开发计算算法， 整合基因型、基因表达和表型数据以构建模型， 描述序列变异如何扰乱调节网络，改变信号处理 并表现在细胞表型上。 我们的方法基于贝叶斯网络，这是我们为 从高通量数据重建分子网络。我们最近应用了这个 框架来开发Geronemo算法，我们将该算法应用于酵母，并发现了一个 序列特异性RNA因子PUF3和P-体之间的新关系 作为MKT1中的单核苷酸多态(SNP)，调节这种关系。 这两项新发现在发现后都得到了实验验证。 我们的方法是基于基因序列和基因序列之间的互补对偶性 功能基因组学。FINE对表型有显著的影响 调整了对控制细胞活动的复杂调控网络的扰动。 单基因表达的变化更容易处理，可以作为一种 帮助将遗传因素与更复杂的下游变化相关联的中介 以层次分明的方式表现。相反，DNA序列的多态是 有效的扰动抗原，提供丰富的变异来源，帮助揭示调控 分子网络中的关系以及它们之间的因果关系。 我们将使用大量高度可变的酵母菌株来开发我们的方法， 我们已经在许多情况下生成了强劲的数量增长曲线 环境条件。在酵母中试行的方法将扩展到基因分型 和来自肿瘤样本的基因表达数据，试图阐明 驱动它们增殖的遗传因素。这些工具将公之于众， 包括友好的图形用户界面和可视化。

The focus of the proposed research is to understand the effect of sequence variation on the function of molecular networks. We will develop computational algorithms that integrate genotype, gene expression and phenotype data to construct models that describe how sequence variation perturbs the regulatory network, alters signal processing and is manifested in cellular phenotypes. Our approach is based on Bayesian networks, a framework we pioneered for the reconstruction of molecular networks from high-throughput data. We recently applied this framework to develop the Geronemo algorithm which we applied to yeast and uncovered a novel relationship between the sequence specific RNA factor PUF3 and P-Bodies, as well as a Single Nucleotide Polymorphism (SNP) in MKT1 that modulates this relationship. Both novel findings were experimentally validated subsequent to their discovery. Our approach is based on the complementary duality between genetic sequence and functional genomics. A significant influence of genotype on phenotype is induced by fine tuned perturbations to the complex regulatory network that governs a cell's activity. Variation in the expression of a single gene is more tractable and can be used as an intermediary to help associate genetic factors to the more complex downstream changes in phenotype in a hierarchical fashion. Conversely, DNA sequence polymorphisms are effective perturb-agens which provide a rich source of variation to help uncover regulatory relations in the molecular network as well as direct their causality. We will develop our methods using a large collection of highly variable yeast strains, for which we have generated robust quantitative growth curves under numerous environmental conditions. The methodologies piloted in yeast will be extended to genotype and gene expression data derived from tumor samples to attempt to elucidate the multiple genetic factors that drive their proliferation. These tools will be made publicly available, including a friendly graphical user interface and visualization.