MOIR - Machine Learning and Modeling Core

项目总结 尽管抗逆转录病毒疗法持久地抑制了病毒复制，但艾滋病毒在感染者中无限期地存在。 治愈HIV-1感染的几种有希望的途径已经曝光，包括CCR5的基因编辑 CD4T细胞受体、抗PD1单抗免疫治疗及广谱中和输注 抗体。这些治疗方法构成了广泛了解 与建立和维护艾滋病毒储存库有关的基本机制，这将 最终为未来更精细的治疗确定关键的新靶点。此外，中的异构性 人类的免疫功能已被映射到多种环境因素，如微生物组、代谢组 和饮食，其中一些与维持艾滋病毒-1宿主有关。在这篇P01中，我们 假设特定的关键代谢物、微生物和其他环境因素影响 对针对艾滋病毒宿主的不同治疗方法的反应性。该计划的主要目标是 机器学习和建模核心(MLMC)将汇集由项目生成的所有数据集 1-3成为一个有凝聚力的整体，以生成艾滋病毒储存库维护的机械模型。我们应该去看看 了解代谢物如何调节免疫转录组和许多亚组的表观基因组。在目标1中，MLMC 将为所有项目提供统计和生物信息学支持，并确定艾滋病毒反弹的关键相关因素 而艾滋病毒DNA从所有大规模数据集(组学)中腐烂。在目标2中，MLMC将进行综合分析 使用在目标1中产生的新数据集。在目标3中，核心将整合监管的并行模型 将艾滋病毒宿主纳入全球统一模型，其中关键的经常性特征将被确定为未来的主要目标 治疗途径。因此，MLMC将作为这一U19的中心资源，以整合所有 数据集和机械洞察力的生成。

PROJECT SUMMARY Despite the durable suppression of viral replication by ART, HIV persists indefinitely in infected individuals. Several promising avenues to cure HIV-1 infection have come to light, including gene editing of the CCR5 receptor in CD4 T cells, anti-PD1 monoclonal antibody immunotherapy and the infusion of broadly neutralizing antibodies. These therapeutic approaches constitute a prime opportunity to extensively understand the underlying mechanisms associated with the establishment and maintenance of the HIV reservoir, which will ultimately serve to identify key novel targets for future more refined therapies. Furthermore, the heterogeneity in human immune function has been mapped to multiple environmental factors such as microbiome, metabolome and diet, some of which have been associated with the maintenance of the HIV-1 reservoir. In this P01, we hypothesize that specific key metabolites, microbes and other environmental factors influence the responsiveness to different therapeutic approaches targeting the HIV reservoir. The main objective of the Machine Learning and Modeling Core (MLMC) will be to bring together all datasets generated by Projects 1-3 into a cohesive whole to generate mechanistic models of HIV reservoir maintenance. We shall look into how metabolites modulate the immune transcriptome and epigenome of many subsets. In Aim 1, the MLMC will provide statistical and bioinformatics support for all projects and identify key correlates of HIV viral rebound and HIV DNA decay from all large-scale datasets (OMICs). In Aim 2, the MLMC will perform integrative analysis using novel datasets generated in Aim 1. In Aim 3, the Core will integrate parallel models of regulation of the HIV reservoir into a global unified model where key recurrent features will be identified as prime targets for future therapeutic avenues. The MLMC will thus serve as the central resource for this U19 for the integration of all datasets and generation of mechanistic insights.