Integrative Approach to Characterizing Gene Regulation

描述(由申请人提供)： 微阵列实验是在系统尺度上分析基因表达水平的有力方法。微阵列技术可以阐明在病理或应激条件下基因表达是如何改变的，并提供对疾病分子机制的洞察。然而，在一些微阵列实验中，基因内点对点的变异性比人们从稳定性考虑的预期要大得多，这意味着为了更好地解释实验结果和改进微阵列设计，有必要更好地理解核酸杂交的动力学。 微阵列使研究人员能够快速获得数千个基因同时表达水平的定量数据。然而，要确定这些数据相对于关于基因、基因产品、组织、细胞和生物体的大量科学信息的重要性，就需要应用统计技术。由于对这些问题的兴趣和许多研究人员的共同努力，现在有几种不同的数据分析技术可用。然而，现有的方法往往忽略了数据中固有的不确定性以及它们对不同基因表达水平之间的交叉相关性估计的影响。 鉴于这些挑战，拟议研究的主要目标是：(I)对微阵列中的DNA杂交有一个基本的了解，(Ii)开发能够区分基因表达水平变化之间的真实相关性和由于噪声而出现的虚假相关性的算法。为了实现第一个目标，我们将开发一个新的微阵列DNA杂交的介观模型。为了实现第二个目标，我们将推广随机矩阵理论方法来研究不同基因表达水平变化之间的交叉相关性。 对这些问题的更好理解将导致开发更准确的工具来研究基因调控网络中的信息交换，使人们能够更好地预测扰动对细胞状态的影响。我们的目标也有可能导致对控制多细胞发育的机制的更深入的理解，并阐明病理细胞事件，包括人类疾病的发生和发展。此外，我们的研究将有助于更好地评估药物对疾病进展患者的效果，并简化预先区分正常、携带者和疾病基因型的过程。

DESCRIPTION (provided by applicant): Microarray experiments are a powerful method for the analysis of gene expression levels at a systems scale. Microarray techniques can illuminate how gene expression is modified under pathological or stressful conditions, and provide insight into the molecular mechanisms of disease. However, intra-gene spot-to-spot variability in some microarray experiments is much larger than one would expect from stability considerations, implying that a better understanding of the kinetics of nucleic acid hybridization is necessary in order to better interpret experimental results and to improve microarray design. Microarrays enable researchers to quickly obtain quantitative data for the simultaneous expression levels of thousands of genes. However, the determination of the significance of such data vis-a-vis the vast amounts of scientific information available on genes, gene products, tissues, cells and organisms, requires the application of statistical techniques. Thanks to the interest in these problems and the concerted effort of many researchers, several different techniques for data analysis are nowadays available. However, available methods often disregard the inherent uncertainties in the data and their effect on the estimation of cross correlations among expression levels of different genes. In view of these challenges, the main goals of the proposed research are: (i) to obtain a fundamental understanding of DNA hybridization in microarrays, and (ii) to develop algorithms that are able to distinguish true correlations between changes in expression levels of genes from spurious correlations that appear due to noise. To achieve the first goal, we will develop a new meso-scale model for DNA hybridization in microarrays. To achieve the second goal, we will generalize random matrix theory methods to the study of cross-correlations among changes in expression levels for different genes. An improved understanding of these questions will lead to the development of more accurate tools for the study of information exchange within gene regulation networks, enabling one to better predict the effect of perturbations to the state of a cell. Our goals also hold the potential to lead to a deeper understanding of the mechanisms that control multi-cellular development and to shed light on pathological cellular events, including the onset and progression of human disease. Moreover, our research will help to better assess the effect of drugs on patients undergoing disease progression, and ease the process of distinguishing normal, carrier and disease genotypes beforehand.