Using three-dimensional protein networks to uncover immuno-modulatory molecular phenotypes in infectious disease

点击翻译按钮获取中文摘要

Using three-dimensional protein networks to uncover immuno-modulatory molecular phenotypes in infectious disease CHALLENGE: Over the past decade, technologies for deep profiling of the human immune system, both in the context of natural and vaccine-mediated immunity, have become readily available. These approaches have generated a wide range of molecular profiles across infectious disease contexts. However, existing studies primarily focus on individual `omic datasets, and do not take into account the underlying molecular networks. Thus, the primary emphasis has been on uncovering predictive biomarkers, but these biomarkers may often be correlative surrogates and have little or no connection with the underlying molecular phenotypes driving disease pathophysiology. GOAL I propose to develop and use a novel framework to integrate genomic data with three-dimensional (3D) structurally-resolved protein networks to uncover immuno-modulatory molecular phenotypes in infectious disease. While protein networks are typically viewed as two-dimensional, with proteins as nodes and interactions between them as edges, this simplifying representation fails to take into account the 3D structures of the proteins themselves, and the corresponding interaction interfaces. My past work has demonstrated the critical importance of taking into account corresponding structural information in the integration of Mendelian mutations with protein networks, to elucidate molecular phenotypes underlying the corresponding genetic disorders, with high sensitivity and specificity. Here, I propose to develop a novel framework that integrates structural genomic data with host-pathogen protein interactome networks to generate 3D host-pathogen interactomes. These 3D interactome networks are then integrated with host (human) genetic data to uncover immuno-modulatory molecular phenotypes in HIV and influenza. INNOVATION AND IMPACT: The proposed work integrates both two orthogonal facets of my expertise in network systems biology and machine learning, and pushes the envelope on multiple key frontiers. First, it provides a novel framework for the integration of host genetic data with host-pathogen protein networks. Second, a key novelty is the incorporation of structural information corresponding to host-pathogen protein interaction interfaces to refine the traditional principle of “guilt-by-association”, and hone in on specific molecular phenotypes that modulate infectious disease risk. The identified molecular phenotypes will generate key mechanistic hypotheses regarding corresponding disease pathophysiology, and help design interventional strategies. Finally, while the focus here is to use this approach in HIV and influenza, the framework itself is generalizable and can be used across infectious disease contexts.