RAPID: Identifying Biophysical Determinants of Binding to the SARS-CoV-2 Main Viral Protease

斯隆·凯特琳癌症研究所的约翰·乔德拉（John Chodera）因确定抑制SARS-CoV-2主要病毒蛋白酶（Mpro）的生物物理决定因素而获得化学系化学理论、模型和计算方法项目的奖励。Mpro是导致COVID-19的病毒中的一种必需酶。Chodera实验室开发了由廉价的消费级图形处理单元（gpu）加速的物理模型，以预测哪些小分子可能结合并抑制与疾病相关的蛋白质。Chodera实验室是Folding@home联盟的一部分，这是一个研究合作，使用Folding@home分布式计算环境来运行这些计算。这个计算资源是由世界各地的志愿者网络捐赠的。最近，为应对COVID-19大流行，Folding@home成为世界上最大的计算平台，拥有2500万个CPU内核和600K gpu随时参与。Chodera实验室将使用Folding@home整合计算和实验，以快速识别Mpro的高亲和力抑制剂，并阐明有效抑制所需的关键相互作用。他们与Informatics Matters（一个致力于枚举合成可行化合物的团队）、Enamine（合成化合物）、Diamond Light Source（使化合物结晶）、Weizmann伦敦实验室（测定化合物）和PostEra（以一种可以加速其他研究实验室和制药公司对Mpro抑制的研究的方式快速和公开地提供结果）的合作者合作。在过去的两个月里，Folding@home已经成为世界上最大的计算资源（每秒2.5百亿亿次浮点运算，2500万个CPU内核，600K gpu），为COVID-19特定研究服务。Folding@home联盟的创始人之一John Chodera和合作者已经建立了一个快速的管道，从14B化合物Enamine REAL Space虚拟合成库中的分子选择到关键的生物物理数据（x射线结构和亲和力）到SARS-CoV-2主要病毒蛋白酶（Mpro），周转时间约为2周。他的实验室现在正在使用相对炼金术自由能方法来评估从最初筛选弱抑制剂到高亲和力配体的68个小分子x射线结构的初始组的快速进展策略。该团队还确定了Mpro活性位点内高亲和力配体结合的关键生物物理决定因素，并对小分子力场的未来准确性进行了基准测试，以指导下一代力场的开发。该实验室正在从Enamine REAL Space中选择分子进行合成，通过DiamondMX/XChem浸泡产生x射线结构，并通过已经存在的合作，由Weizmann研究所的伦敦实验室进行Mpro抑制试验。所有计算数据正在通过美国国家科学基金会资助的分子科学软件研究所（MolSSI） COVID-19分子结构和治疗中心以及开放科学COVID登月计划在线快速传播，以最大限度地发挥基础研究的多种下游用途，并有机会在应用和转化领域产生更广泛的影响。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

John Chodera of the Sloan Kettering Institute for Cancer Research is supported by an award from the Chemical Theory, Models and Computational Methods program in the Division of Chemistry to identify the biophysical determinants of inhibition for the SARS-CoV-2 main viral protease (Mpro). Mpro is an essential enzyme in the virus that causes COVID-19. The Chodera lab develops physical models accelerated by inexpensive consumer-grade graphics processing units (GPUs) to predict which small molecules might bind and inhibit disease-relevant proteins. The Chodera lab is part of the Folding@home Consortium, a research collaboration that uses the Folding@home distributed computing environment to run these calculations. This computing resource is donated by a network of volunteers around the world. Recently, in response to the COVID-19 pandemic, Folding@home became the largest computing platform of any kind in the world, with 25M CPU cores and 600K GPUs participating at any given time. The Chodera lab will use Folding@home to integrate computation and experiment to rapidly identify high-affinity inhibitors of Mpro and to elucidate key interactions required for effective inhibition. They work with collaborators at Informatics Matters (a team that works to enumerates synthetically feasible compounds), Enamine (to synthesize compounds), the Diamond Light Source (to crystallize chemical compounds), the London lab at the Weizmann (to assay compounds), and PostEra (to make the results rapidly and publicly available in a manner that can accelerate research on Mpro inhibition in other research laboratories and pharmaceutical companies). Over the last two months, Folding@home has become the world’s largest computing resource (2.5 exaflops, 25M CPU cores, 600K GPUs) in service of COVID-19 specific research. John Chodera, a founding investigator in the Folding@home Consortium, and collaborators have established a rapid pipeline to go from the selection of molecules within the 14B compound Enamine REAL Space virtual synthetic library to key biophysical data (X-ray structures and affinities) to SARS-CoV-2 main viral protease (Mpro) with ~2 week turnaround time. His laboratory is now using relative alchemical free energy methods to assess strategies for rapidly progressing an initial set of 68 small molecule X-ray structures from an initial screen for weak inhibitors toward high-affinity ligands. The team also identifies key biophysical determinants of high-affinity ligand binding within the active site of Mpro, and benchmark the propsective accuracy of small molecule force fields to inform the development of next-generation force fields. The laboratory is selecting molecules from Enamine REAL Space to be synthesized, soaked to produce X-ray structures by DiamondMX/XChem, and assayed for Mpro inhibition by the London lab at the Weizmann Institute via collaborations already in place. All computational data is being rapidly disseminated online via the NSF-funded Molecular Sciences Software Institute (MolSSI) COVID-19 Molecular Structures and Therapeutics Hub and the open science COVID Moonshot program to maximize multiple downstream uses for fundamental research and the opportunity for broader impacts in applied and translational areas.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.