Discovery and Optimization of Enantioselective Catalysts Guided by Informatics and Machine Learning

在NSF化学部化学催化计划的支持下，伊利诺伊大学香槟分校(UIUC)化学系的Scott E.Denmark教授和他的团队正在努力开发一种新的催化过程设计方法，将以多样性为导向的合成的创造力与信息学的计算能力结合起来。该计划的基本要求是发明和实现高分辨率的化学描述符，能够以计算可读的形式准确地反映分子的化学性质。通过为实验者提供更好的工具来设计具有理想的化学特性的新化合物，从靶向/治疗选择性到光学和材料特性，开发一种真正通用的、通过计算指导分子功能优化的工作流程可能会产生广泛的科学影响。这里最令人感兴趣的是发现和优化用于工业相关化学过程的催化剂。所计划的研究活动是在准备有机化学和数据科学的界面上对学生进行智力和实践训练的理想选择。这一活动的统一主题是发明挑战当前思维的新化学反应。在教育外展方面，要求提供资金，以支持大学圣埃尔莫·布雷迪暑期研究学者计划赞助的三年中每年一名本科生实习生。这个项目是一个为期10周的暑期研究体验，旨在增加来自科学服务不足群体的学生的比例。丹麦教授和他的UIUC团队将以迭代的方式开发一项将计算分析与实验相结合的研究计划，即：(1)在基于给定支架的假设催化剂结构的大型库的计算机生成中，随后计算每个库成员的描述符，(2)多样性分析以生成具有代表性的“训练集”，(3)训练集的合成，(4)在给定反应中对训练集的评估，(5)开发和验证将经验输出与分子性质相关联的数学模型，(6)将该模型应用于催化剂的虚拟库，(7)最佳预测催化剂的合成和评价，以及(8)重复步骤4-7，直到获得所需的产量。这种方法最强大的特征之一是，训练集适用于各种反应，这些反应对支架的催化很敏感。因此，所有三个目标都包括已建立的化学信息学/实验工作流程。在每个目标中，将评估为每个支架生成的训练集，以优化几个化学反应。由于催化剂选择过程仅使用催化剂描述符来执行，因此该过程与反应和机理无关，并且所选择的催化剂训练集可用于评估许多转化。每个目标都涉及配体或催化剂的生成、合成、评估和优化：目标1-用于反式选择性环卤化反应的Brnsted酸催化剂；目标2-对映选择性氧化亚硝烯反应和还原Heck反应；以及目标3-有机金属催化下的对映选择性环氧化物和氮杂环丙烷去对称化。该奖项反映了NSF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。

With support from the Chemical Catalysis Program in the NSF Division of Chemistry, Professor Scott E. Denmark of the Department of Chemistry at the University of Illinois, Urbana-Champaign (UIUC) and his team are working to develop a new approach to the design of catalytic processes that combines the creative power of diversity-oriented synthesis with the computational power of informatics. The foundational requirements for this program are the invention and implementation of high-resolution chemical descriptors that are able to accurately reflect the chemical properties of molecules in computationally readable form. The development of a truly general, computationally guided workflow for the optimization of molecular function could have broad scientific impact by providing the experimentalist with better tools to design new compounds with desirable chemical properties from target/therapeutic selectivity to optical and materials properties. Of paramount interest here is the discovery and optimization of catalysts for industrially relevant chemical processes. The planned research activities planned are ideal for the intellectual and practical training of students at the interface of preparative organic chemistry and data science. The unifying theme of this activity is the invention of new chemical reactions that challenge current thinking. In the context of educational outreach, funds are requested to support an undergraduate intern each of the three years under the auspices of the St. Elmo Brady Summer Research Scholars Program at UIUC. This program is a 10-week summer research experience to increase the percentages of students from underserved groups in the sciences. Professor Denmark and his UIUC team are will developing a research program that combines computational analysis with experimentation, in an iterative fashion, namely: (1) in silico generation of a large library of hypothetical catalyst structures, based on a given scaffold, followed by calculation of descriptors of each library member, (2) diversity analysis to generate a representative “training set”, (3) synthesis of the training set, (4) evaluation of the training set in a given reaction, (5) development and validation of a mathematical model of that correlates empirical output with molecular properties (6) application of that model to the virtual library of catalysts, (7) synthesis and evaluation of best predicted catalysts, and (8) the repetition of steps 4-7 until desired output is achieved. One of the most powerful features of this approach is that the training set is applicable to a wide variety of reactions that are susceptible to catalysis by that scaffold. Accordingly, all three objectives encompass the established chemoinformatic/experimental workflow. Within each objective, the training set generated for each scaffold will be evaluated for the optimization of several chemical reactions. Because the catalyst selection process is performed using catalyst descriptors alone, this process is reaction- and mechanism-agnostic, and chosen catalyst training sets can be used to evaluate many transformations. Each objective involves the generation, synthesis, evaluation and optimization a ligand or catalyst: Objective 1 – Brønsted acid catalysts for atropo-selective cyclohalogenation; Objective 2 – Enantioselective oxidative nitroso-ene reactions and reductive Heck reactions and Objective 3 – Enantioselective epoxide and aziridine desymmetrization under organometallic catalysis.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.