Computational Prediction of Enantioselectivity in Metal-Catalyzed Reactions

在化学系化学催化项目的支持下，圣母大学的Olaf Wiest和Paul Helquist教授正在开发计算方法，用于快速准确地预测化学反应的立体化学或三维结果。在许多反应中，可以形成两种不同的产物，称为对映体或镜像。对于诸如药物的应用，有时只有一种对映异构体是有益的，而另一种可能没有益处或甚至可能具有有害作用。传统的方法完全基于实验研究，以找到一种方法来选择性地产生只有一个对映体是缓慢和昂贵的。它们主要采用试错法，通常需要使用昂贵的试剂和催化剂进行数百次实验，同时产生不必要的化学废物。相比之下，作为当前项目基础的计算方法是预测将产生所需对映体结果的特定催化剂的快速且具有成本效益的手段。这些预测加速并支持后续的实验研究。这项工作可能会产生更广泛的影响，因为它是与一家主要的国际制药公司阿斯利康密切合作进行的。他们不仅将圣母院的方法纳入他们的工艺研究，以促进创新，而且还通过工业实习，在瑞典阿斯利康跨学科培训以及与工业研究人员的定期互动，为从事该项目的学生提供独特的专业发展机会。两位首席研究员坚定地致力于扩大参与，长期以来一直为代表性不足的群体的学生提供本科研究机会，作为高尔文学者项目的导师，为教育背景较差的学生提供指导，以及在为代表性不足的少数民族学生建立桥梁项目中担任导师。圣母大学的赫尔奎斯特和韦斯特教授继续开发了一种用于对映选择性催化（CatVS）的虚拟筛选工具，该工具使用量子引导分子力学（Q2 MM）方法导出的过渡态力场（TSFF）。这包括几种新的优化TSFF的方法，这些方法提高了TSFF与文献力场组合的准确性，使用混合ε约束/PSO方法的半自动拟合过程;改进了虚拟文库（LibGen）生成和筛选的工作流程，与其他程序的新接口，用于构象搜索和TSFF，用于团队与其工业界合作选择的新反应伙伴一旦测试和验证，所有新的开发将被纳入Q2 MM/LibGen/CatVS代码中，该代码将继续在github.com/q2mm上免费提供给科学界。这种方法的影响将被证明在应用范围从Ir催化加氢胺化未活化的烯烃通过新的镍催化的交叉偶联反应的计算机辅助设计的新的配体类的对映选择性合成。计算和实验相结合的方法，再加上与阿斯利康的密切合作，为本科生和研究生提供了一个独特的跨学科培训环境。该奖项反映了NSF的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

With the support of the Chemical Catalysis Program in the Division of Chemistry, Professors Olaf Wiest and Paul Helquist at the University of Notre Dame are developing computational methods for the fast and accurate prediction of the stereochemical, or three-dimensional, outcomes of chemical reactions. In many reactions, two different products known as enantiomers or mirror images can be formed. For applications such as pharmaceuticals, sometimes only one of the enantiomers is beneficial while the other may have no benefit or may even have deleterious effects. Traditional methods based entirely on experimental studies to find a means to selectively generate only one enantiomer are slow and expensive. They make use of largely trial-and-error approaches often requiring hundreds of experiments to be performed with expensive reagents and catalysts and, in the meantime, generate unnecessary chemical waste. In contrast, the computational methods that are the basis of the current project are a fast and cost-efficient means to predict specific catalysts that will produce the desired enantiomeric outcome. These predictions accelerate and support follow-up experimental studies. This work will likely have a significant broader impact because it is performed in close collaboration with AstraZeneca, a major international pharmaceutical company. They not only incorporate the Notre Dame methods into their process research to foster innovation, but also provide unique professional development opportunities for the students working on the project through industrial internships, interdisciplinary training at AstraZeneca in Sweden, and regular interactions with industrial researchers. The two lead investigators are strongly committed to broadening participation with a long track record of offering undergraduate research opportunities to students from underrepresented groups, as instructors in the Galvin Scholars Program for students from less well-prepared educational backgrounds, and by serving as mentors in the Building Bridges Program for underrepresented minority students.Professors Helquist and Wiest at the University of Notre Dame continue the development of a virtual screening tool for enantioselective catalysis (CatVS) that uses transition state force fields (TSFFs) derived using the quantum-guided molecular mechanics (Q2MM) method. This includes several new methods for the optimization of TSFFs that improve the accuracy of TSFF combinations with literature force fields, the semi-automation of the fitting procedure using a hybrid epsilon-constraint/PSO method; improved workflows for the generation and screening of virtual libraries (LibGen), new interfaces to other programs for conformational searching and TSFFs for new reactions selected by the team in collaboration with their industrial partners. Once tested and validated, all new developments will be incorporated in the Q2MM/LibGen/CatVS code that will continue to be available to the scientific community free of charge at github.com/q2mm. The impact of this method will be demonstrated in applications ranging from Ir-catalyzed hydroaminations of unactivated alkenes through novel nickel-catalyzed cross-coupling reactions to the computer-assisted design of new ligand classes for enantioselective synthesis. The combined computational and experimental approach, combined with the close collaboration with AstraZeneca, provides a unique interdisciplinary training environment for undergraduate and graduate students.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.