Leveraging Language Model Multitasking To Predict C–H Borylation Selectivity
Open Access
- 6 May 2024
- journal article
- research article
- Published by American Chemical Society (ACS) in Journal of Chemical Information and Modeling
- Vol. 64 (10) , 4286-4297
Abstract
C-H borylation is a high-value transformation in the synthesis of lead candidates for the pharmaceutical industry because a wide array of downstream coupling reactions is available. However, predicting its regioselectivity, especially in drug-like molecules that may contain multiple heterocycles, is not a trivial task. Using a data set of borylation reactions from Reaxys, we explored how a language model originally trained on USPTO_500_MT, a broad-scope set of patent data, can be used to predict the C-H borylation reaction product in different modes: product generation and site reactivity classification. Our fine-tuned T5Chem multitask language model can generate the correct product in 79% of cases. It can also classify the reactive aromatic C-H bonds with 95% accuracy and 88% positive predictive value, exceeding purpose-developed graph-based neural networks.Keywords
Funding Information
- Engineering and Physical Sciences Research Council
- Exscientia
This publication has 41 references indexed in Scilit:
- Transformer Performance for Chemical Reactions: Analysis of Different Predictive and Evaluation ScenariosJournal of Chemical Information and Modeling, 2023
- Rings in Clinical Trials and Drugs: Present and FutureJournal of Medicinal Chemistry, 2022
- Unified Deep Learning Model for Multitask Reaction Predictions with ExplanationJournal of Chemical Information and Modeling, 2022
- SMILES Pair Encoding: A Data-Driven Substructure Tokenization Algorithm for Deep LearningJournal of Chemical Information and Modeling, 2021
- Late-Stage FunctionalizationChem, 2020
- Molecular Transformer: A Model for Uncertainty-Calibrated Chemical Reaction PredictionACS Central Science, 2019
- Computational Studies of Synthetically Relevant Homogeneous Organometallic Catalysis Involving Ni, Pd, Ir, and Rh: An Overview of Commonly Employed DFT Methods and Mechanistic InsightsChemical Reviews, 2015
- New aryl/heteroaryl CN bond cross-coupling reactions via arylboronic acid/cupric acetate arylationTetrahedron Letters, 1998
- Synthesis of diaryl ethers through the copper-promoted arylation of phenols with arylboronic acids. An expedient synthesis of thyroxineTetrahedron Letters, 1998
- New N- and O-arylations with phenylboronic acids and cupric acetateTetrahedron Letters, 1998