Statistical and Computational Methods in Genetic Analysis

大型和复杂的遗传数据集对标准数据分析技术提出了很大的挑战。在许多情况下，标准的方法是不可行的，甚至不可能分析这些数据。本研究旨在发展与复杂人类谱系数据分析相关的统计和计算方法。第一个主要重点是解决涉及大型复杂家系、多个多态性标记、不完全基因型和复杂疾病模型的问题。第二个主要重点是进一步研究卡方（CHS）重组模型，开发新的技术，将CHS纳入基因定位的方法，以实现更高的数据效率，并应用这种方法来研究人类基因组中的遗传干扰。在这个项目中的大多数拟议的研究是使用马尔可夫链蒙特卡罗（MCMC）方法进行。迄今为止，MCMC方法在人类谱系分析中的探索表明，这种方法非常适合估计概率和似然性。然而，由于现代人类遗传学的模型和方法的特殊性，需要对标准MCMC方法进行特殊的修改，以使该技术在各种遗传作图问题中有效。本项目继续进行这些修改并探索新的应用。在过去的十年中，人类遗传学领域取得了迅速的进展。庞大而复杂的数据集正在以令人难以置信的速度积累。一些疾病基因（其中大多数是单基因简单遗传病）已经被识别和定位。这包括导致囊性纤维化、亨廷顿氏病和一些乳腺癌的基因。这在遗传咨询、基因检测和筛选、药物发现和遗传治疗方面具有巨大的意义。识别复杂疾病的遗传因素是一项困难得多的任务。复杂疾病可能是由不同的易感基因引起的遗传异质性，也可能是由基因与可能的环境效应的组合引起的。许多常见疾病具有复杂的病因，并且被认为至少部分是由于遗传易感性。常见疾病如糖尿病、酒精依赖和某些形式的癌症都是复杂疾病的例子。这项研究开发的方法可以处理复杂的遗传模型，并充分利用现有的遗传数据，从而提高了绘制复杂疾病基因图谱的能力。

Large and complex genetic data sets present a great deal of challenges to standard data analysis techniques. In many cases, standard methods are infeasible or even impossible for analyzing such data. This researchis to develop statistical and computational methods relevant to the analysisof complex human pedigree data. The first main focus is to solve problems that involve large complex pedigrees, multiple polymorphic markers, incompletegenotypes, and complex disease models. The second main focus is to study further the Chi-square (CHS) recombination models, to develop new techniques to incorporate CHS into methods of gene mapping to achieve greater efficiencyof data, and to apply this methodology to study genetic interference in the human genome. Most of the proposed research in this project is to be carried out using the Markov chain Monte Carlo (MCMC) methodology. Exploration of MCMC methods in human pedigree analysis thus far shows that this methodology is highly suitable for estimating probabilities and likelihoods. However, because of special features of models and methods appropriate for modern human genetics, special modifications to the standard MCMC approach are requiredfor this technique to be effective in a variety of genetic mapping problems.This project continues the work of making these modifications and exploring new applications.The last decade has seen rapid advancements in the field of human genetics. Large and complex data sets are accumulating at an incredible rate. Some disease genes (most of them are for single-gene simple genetic disorders) have been identified and mapped. This includes the genes responsible for cystic fibrosis, Huntington's disease and some breast cancers. This has tremendous implication in genetic counseling, genetic testing and screening, drug discovery, and genetic therapy. Identification of genetic factors for complex diseases is a far more difficult task. Complex diseases may be geneticallyheterogeneous caused by different susceptibility genes, or may be caused bya combination of genes with possible environmental effects. Many common diseases have complex etiology, and are believed to be at least partially due to genetic predisposition. Common diseases such as diabetes, alcohol dependence, and some forms of cancer are examples of complex disorders. Methods developed in this research can handle complex genetic models and make use of available genetic data fully, thereby increasing the power to map genes for complex diseases.