CAREER: Innovation: The Three R's: A Model-Building Toolkit for Rational, Reproducible, and Rigorous Computational Enzymology

酶学是研究催化蛋白质的结构、能量、功能和化学反应的学科。酶的原子尺度计算机建模是数十亿美元研究工作的一部分，该工作有助于新药的设计，有助于研究蛋白质结构和功能，并促进我们对疾病的分子基础的理解。尽管计算酶学得到了广泛的应用和成功，但模型的组成/大小与模拟的准确性之间的定量关系仍然知之甚少，方法的比较几乎是不可能的。这个项目将设计一个自动化的协议，使用合理创建和可重复的模型进行酶的计算研究，其中假设可以通过数据驱动的方法进行严格测试。为了与生物基础设施部保持一致-S专注于通过投资于生物研究资源的开发和增强来增强生物发现的能力，开发的协议将通过网络平台和用户界面提供。此外，该项目将建立一个生物学入门课程的实验室模块，使本科生熟悉蛋白质数据库、酶动力学和计算机建模。规则发现和模型构建自动化将为计算酶学的可重复、合理和严格的方法铺平道路。改进的研究和项目设计标准将首次允许对生化模拟的准确性进行真正的定量评估。这项研究将影响STEM研究界的一大批多学科领域，从学术界和工业界的结构生物学家、药理学家和计算化学家，到下一代生物和生物化学本科生。该项目的中心目标是设计一种自动化的协议，用于使用合理创建和可重复的模型进行酶的计算研究，其中假设可以通过数据驱动的方法进行严格测试。自动化、基于规则的软件工具包(RINRUS，残基相互作用网络残基选择器的缩写)的软件设计将允许使用可重复和合理创建的模型在原子水平上对酶进行计算研究。RINRUS将通过选择蛋白质结构中的关键原子来指导研究工作流程，并将其包括在计算模型中。然后，RINRUS将产生一个庞大的、计算上容易处理和化学上严格的酶模型库，准备好使用分子建模软件包进行生产质量的模拟。自动化项目设计和研究实践的标准化将使生化界能够将重点放在蛋白质结构和功能中影响更大的现象上。通过建立一个基于网络的储存库和讨论论坛，将以前所未有的规模促进社区数据共享和酶模型的校准。这项研究将创造一个根本性的、多学科的转变，因为几个领域的计算和数据科学家对他们的发现与实验观察相一致或不一致的原因获得了更好的量化理解。通过新颖、互动的讲授材料和实验室模块，生物和化学入门课程的本科生将接触到诺贝尔奖获奖研究和方法。这些活动将在STEM入门课程和现实世界的科学研究之间建立联系，以增加学生的吸引力和留存力，特别是在STEM社区中代表性不足的少数族裔本科生中。该项目的结果可以在www.memphis.edu/chem/faculty-deyonker/publications.php.This上找到，该奖项反映了国家科学基金会的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Enzymology is the study of the structure, energetics, function, and chemical reactions of catalytic proteins. Atomic-scale computer modeling of enzymes is part of the multibillion-dollar research effort that aids the design of new pharmaceuticals, helps investigate protein structure and function, and advances our understanding of the molecular basis of disease. Despite the widespread use and success of computational enzymology, quantitative relationships between the composition / size of the models and accuracy of simulations are still poorly understood, and comparison of methodologies is nearly impossible. This project will design an automated protocol for the computational study of enzymes using rationally-created and reproducible models, where hypotheses can be rigorously tested via a data-driven approach. In keeping with the Division of Biological Infrastructure?s focus on empowering biological discovery by investing in the development and enhancement of biological research resources, the developed protocol will be made available via a web platform and user interface. In addition, the project will establish a laboratory module for introductory biology courses to familiarize undergraduate students with the Protein Data Bank, enzyme kinetics, and computer modeling. Rule discovery and model building automation will pave the way to a reproducible, rational, and rigorous approach to computational enzymology. Improved research and project design standards will allow, for the first time, a truly quantitative assessment of accuracy in biochemical simulations. This research will impact a large, multidisciplinary swath of the STEM research community, from structural biologists, pharmacologists, and computational chemists in academia and industry, to the next generation of biology and biochemistry undergraduates. The central goal of this project is to design an automated protocol for the computational study of enzymes using rationally-created and reproducible models, where hypotheses can be rigorously tested via a data-driven approach. Software design of an automated, rules-based software toolkit (RINRUS, short for Residue Interaction Network-based ResidUe Selector) will allow the computational study of enzymes at the atomic-level using reproducible and rationally-created models. RINRUS will guide research workflows by selecting crucial atoms in a protein structure to be included in computational models. RINRUS will then produce an enormous library of computationally tractable and chemically rigorous enzyme models, ready for production-quality simulations using molecular modeling software packages. Automated project design and standardization of research practices will allow the biochemical community to focus on higher-impact phenomena in protein structure and function. Community data sharing and calibration of enzyme models at an unprecedented scale will be facilitated by creation of a web-based repository and discussion forum. This research will create a fundamental, multidisciplinary shift, as computational and data scientists in several domains obtain an improved quantitative understanding of why their findings agree or disagree with experimental observation. Through novel, interactive lecture materials and a laboratory module, undergraduates in introductory biology and chemistry courses will be exposed to Nobel Prize-winning research and methodologies. These activities will forge bonds between introductory STEM courses and real-world scientific research to increase student attraction and retention, especially among underrepresented minority undergraduates in the STEM community. Results of this project can be found at www.memphis.edu/chem/faculty-deyonker/publications.php.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.