Answers are grounded in this paper and cite their sources.
Evaluation shows machine learning models improve yield prediction and classification in amide coupling reactions, indicating potential benefits for synthetic chemistry.
Reaction optimization is a very time- and resource-intensive process, as optimal reaction conditions depend highly on nuances of substrate identity, and require extensive fine-tuning of synthetic conditions to arrive at the highest-yielding conversions. Amide couplings, which comprise nearly forty percent of synthetic transformations performed in a medicinal chemistry setting, present a particularly challenging context for predictive modeling given the diversity of coupling agents and reaction parameters that must be matched to substrate reactivity. Amide coupling reaction data makes it well-suited for an approach involving machine learning, predicting optimal reaction conditions given a particular substrate feature set. Herein, we report a platform for standardizing and filtering open source reaction data from the ORD (Open Reaction Database) and using this machine-readable dataset to train thirteen machine learning models. These include linear, tree-based, kernel method, instance based, neural network, and ensemble architectures, in the yield prediction and classification of coupling agents in amide coupling reactions. While yield prediction remained a difficult task for the models due to the complexity of our reaction data, the models performed with great accuracy when classifying reactions to their ideal coupling agent category, including carbodiimide-based, uronium salt, and phosphonium salt. To further validate this approach, we deployed the classification models on literature reported data not in the ORD database. Our results demonstrate that kernel methods and ensemble-based architectures perform significantly better than other models such as linear or single tree. Additionally, molecular environment features, captured by XYZ coordinates, three-dimensional features, and Morgan Fingerprints around reactive functional groups, boosted model predictivity more than bulk material properties such as molecular weight, LogP, and SMILES.
No takes yet. Share an insight, caveat, or question.
Chalasani et al. (2025) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: