Using Machine Learning to Access Challenging Hydrogenations: A combined theoretical and experimental approach

使用机器学习来实现具有挑战性的氢化：理论和实验相结合的方法

• 批准号: 2602290
• 金额: --
• 依托单位: University of Cambridge
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2021
• 资助国家: 英国
• 起止时间: 2021 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2602290
• 关键词: Using; Machine; Learning; Access; Challenging

This project will use machine learning methodologies to probe the conditions necessary to undertake efficient asymmetric hydrogenation reactions on tetra-substituted carbon- carbon double bonds to form 1,2- contiguous stereocentres. Formation of such novel stereocentres allows for movement into three-dimensional space for access to new molecules and pharmaceuticals. Asymmetric hydrogenations of tetra-substituted carbon-carbon double bonds are challenging particularly in obtaining high enantiomeric-excess values. To solve this problem machine learning techniques will be employed for the prediction of such values.This will be achieved through the development of a dataset from available literature examples of asymmetric hydrogenation reactions. This dataset will be used for supervised machine learning methodologies for the development of predictive models. Reaction screening will also be undertaken in the lab to enrich the dataset and feed in to the machine learning models using a 'feedback loop' approach. This project is half computational and half experimental as such this project is joint supervised with Dr Ruth Webster for experimental research and Dr Matthew Grayson for computational.

该项目将使用机器学习方法来探讨在四取代的碳-碳双键上进行有效的不对称氢化反应以形成1,2-连续立构中心所需的条件。这种新颖的立体中心的形成允许进入三维空间以获得新的分子和药物。四取代碳-碳双键的不对称氢化具有挑战性，特别是在获得高对映体过量值方面。为了解决这个问题，将采用机器学习技术来预测这些值。这将通过根据不对称氢化反应的现有文献示例开发数据集来实现。该数据集将用于用于开发预测模型的监督机器学习方法。实验室还将进行反应筛选，以丰富数据集，并使用“反馈循环”方法输入机器学习模型。该项目一半是计算性的，一半是实验性的，因此该项目由露丝·韦伯斯特博士（Dr Ruth Webster）和马修·格雷森博士（Matthew Grayson）共同监督进行实验研究和计算。