Biomedical Computing and Informatics Strategies for Infectious Disease Research

 描述(申请人提供)：传染病研究的一个重要目标是开发易感性的遗传预测因子。我们在这一努力中的成功将关键取决于可用于理解高维遗传和基因组数据的信息学方法和软件。这项研究计划的目标是开发、评估、分发和支持新的和新的生物医学计算算法和开源软件，以识别临床上重要传染病结局的遗传预测因素的组合。这项应用将针对通过高通量DNA测序确定的日益增长的稀有遗传变异。我们的临床应用将集中在艾滋病毒/艾滋病临床试验中对抗逆转录病毒应答的预测。我们在这里提出了一个高度创新的层次化稀有变体折叠机器(HRVCM)算法，用于识别和折叠跨基因区域的稀有变体组合(AIM1)。然后，我们将把这些新的崩溃的HRVCM变量整合到我们流行的多因素降维(MDR)方法中，该方法将结合来自全基因组关联研究的常见单核苷酸多态(SNP)或GWA(AIM 2)来评估它们。我们新的HRVCM-MDR方法将首次使在传染病遗传学研究中同时评估稀有和常见变异之间的非相加相互作用成为可能。我们将把这些新的和新的方法应用于来自艾滋病临床试验小组(ACTG)研究的近3000名受试者的大约1300万个罕见和常见的变种，该研究旨在评估含有Eefavirenz的抗逆转录病毒疗法(ART)方案(AIM 3)的病毒学失败风险。最后，我们将通过我们免费提供的MDR软件包(AIM4)将所有方法作为开源发布给生物医学研究社区。

 DESCRIPTION (provided by applicant): An important goal of infectious disease research is to develop genetic predictors of susceptibility. Our success in this endeavor will depend critically on the informatics methods and software that are available for making sense of high-dimensional genetic and genomic data. The goal of this research program is to develop, evaluate, distribute and support new and novel biomedical computing algorithms and open-source software for identifying combinations of genetic predictors of clinically important infectious disease outcomes. This application will target the growing body of rare genetic variants identified by high-throughput DNA sequencing. Our clinical application will focus on the prediction of antiretroviral response in clinical trials for HIV/AIDS. We propose here a highly innovative Hierarchical Rare Variant Collapsing Machine (HRVCM) algorithm for identifying and collapsing combinations of rare variants across gene regions (AIM 1). We will then integrate these new collapsed HRVCM variables into our popular Multifactor Dimensionality Reduction (MDR) method that will assess them in combination with common single-nucleotide polymorphisms (SNPs) from genome-wide association studies or GWAS (AIM 2). Our novel HRVCM-MDR approach will, for the first time, make it possible to assess non-additive interactions among sets of rare and common variants simultaneously in genetic studies of infectious diseases. We will apply these new and novel methods to approximately 13 million rare and common variants from nearly 3000 subjects that participated in an AIDS Clinical Trials Group (ACTG) study to evaluate risk for virologic failure with efavirenz-containing antiretroviral therapy (ART) regimens (AIM 3). Finally, we will release all methods as open source to the biomedical research community through our freely available MDR software package (AIM 4).