RAPID: Optimizing Experimental Approaches to Ebola Membrane Fusion Inhibitor Peptide Design through High-Throughput Biomolecular Simulation Workflows on Blue Waters

该项目旨在大幅缩短寻找有前景的抗击埃博拉药物的总体试验时间。埃博拉病毒感染的一个非常早期的阶段是病毒进入，即埃博拉病毒附着并进入宿主细胞。一种很有前途的抗病毒药物是一种“进入抑制”或“进入阻断”的药物。这种药物可以抑制埃博拉病毒进入细胞。为了让这种药物发挥很好的作用，它必须由与埃博拉病毒强结合的埃博拉抑制剂分子组成，也就是说，这些抑制剂必须对病毒具有很强的“结合亲和力”。该项目寻求通过分析有希望的抑制剂，并通过在超级计算机上模拟和优化它们的行为来提高它们的结合亲和力，从而极大地加快寻找埃博拉抑制剂(然后可以加入到抗病毒药物中)的进程。通过“虚拟”测试发现并初步审查抑制剂，只有有希望的抑制剂才会被送往实验室进行测试。这个项目将开发和使用的基于高性能计算的新模拟软件将在Blue Waters Petascale资源上运行。该项目旨在开发可控制的、高通量的模拟工作流程，以促进对埃博拉病毒潜在抑制剂的前导肽的修饰进行快速研究，以提高它们与病毒的结合亲和力。每个实验的结果可以反馈到下一轮的优化中，以加快多肽设计过程。该建议建立在以前所做工作的基础上：(A)从以前的模拟中吸取的经验教训(B)名为Amber MD的模拟包的可用性，该模拟包已经针对Blue Waters上的GPU进行了优化，将使模拟运行得更快，以及(C)用于分析步骤和输入阶段性的原型代码。此外，PI计划使用现有的工作流引擎(如Pegasus)来构建工作流。因此，有一个明确的意图，即制造健壮的软件和重复使用已经存在的软件，这与ACI计划的战略目标一致。

This project seeks to greatly reduce the overall experimentation time for finding promising drugs that fight Ebola. A very early stage of Ebola infection is viral entry, when the Ebola virus attaches to and enters the host cell. A promising type of antiviral drug is one that is "entry-inhibiting" or "entry-blocking". This type of drug inhibits the entry of the Ebola virus into the cell. In order for this drug to work well, it must be made up of Ebola inhibitor molecules that bind strongly to the Ebola virus, that is, these inhibitors must have a strong "binding affinity" to the virus. This project seeks to greatly speed up the process of finding Ebola inhibitors (that can then be incorporated into anti-viral drugs) by analyzing promising inhibitors and improving their binding affinity by simulating and optimizing their behavior on supercomputers. Inhibitors are found and initially vetted through "virtual" testing and only promising inhibitors are sent for laboratory testing. The new software for high-performance computing based simulations, that this project will develop and use, will run on the Blue Waters Petascale Resource. This project seeks to develop steerable, high-throughput simulation workflows that facilitate the rapid investigation of modifications to the lead peptides of potential inhibitors of the Ebola virus to improve their binding affinity to the virus. The results of each experiment can feed back into further rounds of optimization to speed the peptide design process. This proposal builds on work done previously: (a) lessons learned from previous simulations (b) the availability of a simulation package named AMBER MD, that already has code optimized for GPUs on Blue Waters that will make simulations run faster, and (c) prototype code for analysis steps and input staging. Also, the PI plans to use an existing workflow engine (such as PEGASUS) to build the workflows. Thus, there is a stated intent to make robust software and reuse software that already exists, which aligns with the ACI programs strategic goals.