CATH-FunVar - Predicting Viral and Human Variants Affecting COVID-19 Susceptibility and Severity and Repurposing Therapeutics

SARS-CoV-2引起了一场大流行，导致全世界数百万人死亡，并造成重大的社会和经济破坏。尽管疫苗试验令人鼓舞，但疫苗必须在全球范围内分发，在一段时间内还需要治疗性干预。很明显，一些人类群体更容易感染这种疾病。例如，老年男子、黑人和亚裔社区。造成这些差异的因素尚不清楚，虽然社会、经济和文化问题可能很重要，但遗传因素也可能起到作用。此外，引起严重反应和增加发病率的生物学机制尚不清楚。在这个项目中，我们将分析不同人群(如性别、种族、严重反应的人)和SARS-CoV-2中的基因变异(导致蛋白质的驻留突变)。我们将使用结构和进化数据来确定这些突变是否会影响病毒和人类蛋白质之间的结合。突变确实影响结合的人类蛋白质将被映射到蛋白质网络，以确定可能受到影响的生物途径。我们有功能强大的工具来注释蛋白质和它们在其中运作的途径模块。我们的数据将使对疾病严重性的影响合理化，并改善对高危人群的诊断。最后，这些途径中的蛋白质很可能成为有效的药物靶点，我们将使用我们的蛋白质家族数据来识别或重新使用副作用较低的合适药物。我们也会分析相关的冠状病毒，以确定未来的风险。我们已经建立了一个网站(https://funvar.cathdb.info/uniprot/dataset/covid)，提供SARS-CoV-2病毒蛋白质图谱、功能注释和突变与已知/预测的功能位点的接近程度。这是目前填充的初步试点数据。研究计划我们将：(A)将与病毒蛋白相互作用的‘人类相互作用蛋白’分类为Cath-FunFams，提取已知或预测的结构，并将不同性别和人群的变异(残基突变)映射到这些结构上。(B)执行FunVar分析，以确定人类相互作用蛋白和SARS-CoV-2蛋白中可能对功能产生影响的突变。(C)将人类相互作用蛋白映射到蛋白质网络，以突出宿主反应中涉及的生物过程，以及不同性别/种族之间的不同影响(D)识别临床上有影响的人类相互作用因素(E)通过FunVar-COVID19页传播信息我们的管道将在不同的人类群体中检测不同的变异，可能会影响功能，影响新冠肺炎的响应。它还将分析现有的药物数据，以建议可能的治疗方法。此外，我们的流水线将是通用的，还将用于分析可能构成未来风险的其他密切相关的冠状病毒基因组。

SARS-CoV-2 has caused a pandemic resulting in millions of deaths worldwide and significant social and economic disruption. Although vaccine trials have been encouraging vaccines must be distributed globally and therapeutic interventions will be needed for some time. It is clear that some human populations are much more vulnerable to the disease. For example older men and black and Asian communities. The factors causing these differences are still unclear and whilst social, economic and cultural issues are likely to be important, genetic factors could also play a role. Furthermore, the biological mechanisms by which severe responses arise and increase morbidity are still not known.In this project we will analyse genetic variations (causing reside mutations in the proteins) in diverse human populations (e.g. gender, ethnicity, people with severe responses) and in SARS-CoV-2. We will use structural and evolutionary data to determine whether the mutations could affect binding between the virus and human proteins. Human proteins in which mutations do affect binding will be mapped to protein networks to identify biological pathways that could be affected. We have powerful tools for functionally annotating proteins and the pathway modules in which they operate. Our data will rationalise the impacts on disease severity and improve diagnostics for populations at risk. Finally, proteins in these pathways are likely to be effective drug targets and we will use our protein family data to identify or repurpose suitable drugs having low side effects. We will also analyse related coronaviruses to identify future risks.We have already established a website (https://funvar.cathdb.info/uniprot/dataset/covid) providing mapping of SARS-CoV-2 viral proteins, functional annotations and proximity of mutations to known/predicted functional sites. This is currently populated with preliminary pilot data. It will be extended to host interactors and provide information on pathways and repurposed drugs.Research PlanWe will: (a) Classify 'human interactor' proteins interacting with viral proteins into CATH-FunFams to extract known or predicted structures and map variants (residue mutations) from different genders and populations onto these structures.(b) Perform FunVar analyses to identify mutations in human interactor and SARS-CoV-2 proteins likely to have functional impacts.(c) Map human interactors to a protein network to highlight biological processes implicated in host response and differentially affected between different genders/ethnicities(d) Identify human interactors which have clinically approved drugs or which map to FunFams from which clinically approved drugs can be repurposed.(e) Disseminate information via FunVar-COVID19 pages Our pipeline will detect diverse variants in different human populations, likely to be impacting functions and affecting Covid-19 response. It will also analyse available drug data to suggest possible therapeutics. Furthermore, our pipeline will be generic and will also be used to analyse other closely related coronavirus genomes that could pose future risks.