HPMI: Host Pathogen Mapping Initiative

宿主病原体图谱计划 2.0 总体总结 宿主-病原体图谱计划 (HPMI) 2.0 是一项跨学科计划，旨在改善我们的 了解宿主细胞系统和呼吸道病原体之间的相互作用，并建立在 其上一次迭代的成功。最终目标是确定治疗靶点、治疗方式和 预测疾病的严重程度。 HPMI 2.0将重点针对细菌病原体结核分枝杆菌 (Mtb) 和病毒性呼吸道病原体，例如 SARS-CoV-2 和令人关注的变种，例如omicron 和 delta，如 以及流感、副流感病毒和呼吸道合胞病毒等 RNA 病毒。我们将使用 蛋白质组学、遗传学和结构生物学方法来研究与这些传染性相关的宿主因素 疾病相关细胞模型和患者样本中的疾病，并将我们的数据与现有的 -omic 相结合 数据集。将使用网络和结构建模方法来集成这些数据集以使其可测试 预测宿主中调节感染以及疾病预后的蛋白质、复合物和途径。 我们的提案将集中于对人类样本进行分析，以破译潜在传染性的网络 呼吸道疾病。为了生成病毒性呼吸道疾病的临床相关数据集，我们将使用人类 用于系统生物学分析的肺原代细胞和三维人气道类器官（HAO）， 确定 SARS-CoV-2 感染患者血浆样本的整体蛋白质组，并整合我们的数据 与临床数据集。为了揭示结核病易感性异质性的决定因素，我们将利用 结核病易感基因与全基因组序列相结合的全基因组关联研究（GWAS） 牛津大学临床研究中心越南结核病患者队列中感染 Mtb 菌株的研究 （奥克鲁）。我们将进一步分析健康人的肺泡巨噬细胞和外周血来源的巨噬细胞 捐赠者了解针对 Mtb 感染的先天免疫反应机制和细胞类型的具体特征。 我们假设不同的结核病临床结果是由感染者的特定分子网络调节的 巨噬细胞。我们用于识别组合生物标志物以预测疾病结果的模型将是 对越南 Mtb 感染队列和 UCSF 的 COVID-19 患者进行了评估。最后，我们将探索潜力 Mtb 和 SARS-CoV-2 之间的共同点，因为严重形式的 TB 和 COVID-19 都伴随着 加剧炎症反应和“细胞因子风暴”。 对宿主-病原体复合物的功能和机制的更好的分子理解可能会揭示新的 干预治疗策略，包括规避宿主定向治疗的策略 当前使用抗生素或抗病毒药物治疗方案的局限性，其中病原体蛋白的突变可能 降低药物疗效。我们的宿主-病原体细胞图谱将有助于构建可解释的深度学习系统 用于预测传染病风险、疾病严重程度和组合的最先进的建模方法 临床环境中结核分枝杆菌和 SARS-CoV-2 感染的危险因素。

THE HOST PATHOGEN MAP INITIATIVE 2.0 OVERALL SUMMARY The Host-Pathogen Map Initiative (HPMI) 2.0 is an interdisciplinary program that aims to improve our understanding of the interactions between host cellular systems and respiratory pathogens and builds upon the success of its previous iteration. The ultimate goals are identifying therapeutic targets, treatment modalities, and predicting disease severity. HPMI 2.0 will focus its efforts on the bacterial pathogen Mycobacterium tuberculosis (Mtb) and viral respiratory pathogens such as SARS-CoV-2 and variants of concern, e.g. omicron and delta, as well as influenza, parainfluenza virus, and respiratory syncytial virus, among other RNA viruses. We will use proteomics, genetics and structural biology approaches to study the host factors relevant to these infectious diseases in disease-relevant cell models and patient samples, and combine our data with existing -omic datasets. Network and structure modeling approaches will be used to integrate these datasets to make testable predictions about proteins, complexes and pathways in the host regulating infection as well as disease prognosis. Our proposal will be centered on the profiling of human samples to decipher networks underlying infectious respiratory diseases. To generate clinically relevant datasets on viral respiratory disease, we will use human lung primary cells and three-dimensional human airway organoids (HAO) for systems biology analyses, determine the global proteome of plasma samples from SARS-CoV-2-infected patients, and integrate our data with clinical datasets. To uncover the determinants of the heterogeneity of susceptibility to TB, we will utilize genome-wide association studies (GWAS) for TB susceptibility genes combined with whole genome sequences of the infecting Mtb strain in a cohort of TB patients in Vietnam at the Oxford University Clinical Research Unit (OUCRU). We will further profile alveolar macrophages and peripheral blood derived macrophages from healthy donors to understand mechanisms of innate immune responses to Mtb infection and cell-type specific features. We hypothesize that different TB clinical outcomes are regulated by specific molecular networks in infected macrophages. Our models for the identification of combinatorial biomarkers to predict disease outcomes will be assessed on the Mtb-infected cohort in Vietnam and COVID-19 patients at UCSF. Lastly, we will explore potential commonalities between Mtb and SARS-CoV-2, as severe forms of TB and COVID-19 are accompanied by exacerbated inflammatory responses and “cytokine storms”. A better molecular understanding of the functions and mechanisms of host-pathogen complexes may reveal new therapeutic strategies for intervention, including strategies of host-directed therapies that circumvent the limitations of current drug regimens using antibiotics or antivirals where mutations in the pathogen proteins can diminish drug efficacy. Our host-pathogen cell maps will help build interpretable deep learning systems using state-of-the-art modeling approaches for prediction of infectious disease risk, disease severity and combinatorial risk factors for M. tuberculosis and SARS-CoV-2 infection in clinical settings.