Statistical methods for genetic data

描述（由申请人提供）：遗传学在许多人类疾病中起作用，无论疾病本身是遗传性的还是与基因活性的实质性变化有关。通过利用最近开发的技术，使人们能够在全基因组水平上进行生物学研究，存在着更好地了解和诊断人类疾病的巨大机会。有一个实质性的需要，开发新的定量工具，专门设计用于分析这些研究产生的大量数据。拟议研究的总体目标是开发有助于理解基因组数据的统计方法和软件。 特别关注的是功能基因组学，其中来自基因表达阵列和大规模基因分型的数据可用于研究大量基因如何工作以完成各种功能角色。将开发用于DNA微阵列实验的统计推断技术，特别是识别在两种或更多种生物条件下差异表达的基因。这些技术将适用于静态实验和时间过程实验。还将开发用于转录调节的遗传解剖的统计方法。这包括在全基因组和基因特异性水平上估计基因表达的遗传控制的方法，以及绘制显示与基因表达连锁的基因座的方法。特别是，多位点连锁分析模型选择方法将进行调查，将开发新的方法，计算效率高的模型生成，选择和显着性分析。 所有这些方法都将被实现为用户友好的软件，并将免费分发给学术界。这些方法还将与实验人员合作，在公开的数据上进行测试，以验证这些方法提供的结果是否具有生物学意义。总体而言，这项工作的目的是通过为基因组研究提供严格的分析工具，促进对人类疾病的分子生物学和遗传基础的理解。

DESCRIPTION (provided by applicant): Genetics plays a role in many human diseases, whether the disease itself is inherited or it is associated with a substantial change in the activity of genes. A great opportunity exists to better understand and diagnose human disease by utilizing recently developed technologies that allow one to carry out biological studies at the genome-wide level. There is a substantial need to develop new quantitative tools specifically designed to analyze the enormous amounts of data generated by these studies. The overall goal of the proposed research is to develop statistical methods and software useful in understanding genomic data. The particular focus is in functional genomic, where data from gene expression arrays and large-scale genotyping can be used to study how large numbers of genes work to accomplish various functional roles. Statistical inference techniques for DNA micro array experiments will be developed, specifically identifying genes that are differentially expressed among two or more biological conditions. These techniques will be applicable to both static experiments and time course experiments. Statistical methods for the genetic dissection of transcriptional regulation will also be developed. This includes methods to estimate the genetic control of gene expression at both genome-wide and gene-specific levels, and methods to map loci showing linkage to gene expression. In particular, multiple locus linkage analysis from a model selection approach will be investigated, where new methods will be developed for computationally efficient model generation, selection, and significance analysis. All of these methods will be implemented into user-friendly software that will be freely distributed to the academic community. The methods will also be tested on publicly available data in collaboration with experimentalists, in an effort to verify that the methods provide biologically meaningful results. Overall, this work is aimed at contributing to the understanding of the molecular biology and genetic basis of human disease by providing rigorous analytical tools for genomic studies.