Modeling Core

点击翻译按钮获取中文摘要

Abstract – Modeling Core The Modeling Core will integrate and mine heterogeneous multiomics data generated in Projects 1 and 2 and the Technology Core to construct multi-scale models of regulatory and metabolic networks that are causally and mechanistically associated with disease progression and treatment outcomes. In Project 1, we will use the Systems Genetics Network AnaLysis (SYGNAL) pipeline to conduct joint modeling of innate and adaptive immune cell subpopulations from blood samples of human TB progressors, as well as orthologous cell subpopulations from mouse model of human TB progression. As input for model construction, we will use transcriptional, cytokine, chemokine and eicosanoid profiles collected over the course of the disease from disease-relevant immune cell types and tissues (lung and blood). Tractability of the mouse model will help to dissect gene networks and mechanisms underlying the etiology of the disease in the lung and how it relates to predictive signature in the blood. We will use interactions deciphered using the SYGNAL network to generate tissue-specific probabilistic Boolean network (PBN) models. Actionable predictions from SYGNAL and PBN network models will drive experiments to identify genetic perturbations that push the immune response towards desirable states. Using comparative network analysis we will then translate this mechanistic understanding from mouse to orthologous mechanisms in human to make predictive blood signatures actionable in terms of guiding preventive or treatment interventions. The goal of the Modeling Core in Project 2 is to decipher how genetic differences across different strains of Mycobacterium tuberculosis (MTB) alter regulatory and metabolic network responses to generate vastly difference treatment outcomes. The input data for modeling will include transcriptomics (RNA-seq), P-P and P-DNA interactions (ChIP-seq, MS-proteomics), TRIP screens, and metabolomics from bulk and sorted drug-tolerant and persister sub-populations of the four MTB strains, subjected to different drugs and stressors. Using a diverse suite of algorithms, we will mine these multi-omic data to generate Environment and Gene Regulatory Influence Network (EGRIN) models and IntegrateD models for REgulation And Metabolism (IDREAM). We will use these network models to drive experimentation and dissect how genetic variation across MTB strains alters their regulatory and metabolic networks to manifest in vastly different clinical outcomes. Finally, the Modeling Core will work with the Data Management and Bioinformatics Core to make data and models available for exploration, allowing biologists to formulate testable hypotheses.