Journal ArticleDOI
Scaffold‐directed face selectivity Machine‐Learned from vectors of non‐covalent interactions
TLDR
In this article, a method to vectorize and machine-learn non-covalent interactions responsible for scaffold-directed reactions important in synthetic chemistry is described, and models trained on this representation predict correct face of approach in ca. 90% of Michael additions or Diels-Alder cycloadditions.Abstract:
This work describes a method to vectorize and Machine-Learn, ML, non-covalent interactions responsible for scaffold-directed reactions important in synthetic chemistry. Models trained on this representation predict correct face of approach in ca. 90 % of Michael additions or Diels-Alder cycloadditions. These accuracies are significantly higher than those based on traditional ML descriptors, energetic calculations, or intuition of experienced synthetic chemists. Our results also emphasize the importance of ML models being provided with relevant mechanistic knowledge; without such knowledge, these models cannot easily "transfer-learn" and extrapolate to previously unseen reaction mechanisms.read more
Citations
More filters
Journal ArticleDOI
Reaction performance prediction with an extrapolative and interpretable graph model based on chemical knowledge
TL;DR: In this article , a knowledge-based graph model that embeds the digitalized steric and electronic information was developed to enable the learning of the synergistic influence of reaction components.
References
More filters
Journal ArticleDOI
Extended-Connectivity Fingerprints
David Rogers,Mathew Hahn +1 more
TL;DR: A description of their implementation has not previously been presented in the literature, and ECFPs can be very rapidly calculated and can represent an essentially infinite number of different molecular features.
BookDOI
Learning to learn
Sebastian Thrun,Lorien Pratt +1 more
TL;DR: This chapter discusses Reinforcement Learning with Self-Modifying Policies J. Schmidhuber, et al., and theoretical Models of Learning to Learn J. Baxter, a first step towards Continual Learning.
Journal ArticleDOI
Planning chemical syntheses with deep neural networks and symbolic AI
TL;DR: This work combines Monte Carlo tree search with an expansion policy network that guides the search, and a filter network to pre-select the most promising retrosynthetic steps that solve for almost twice as many molecules, thirty times faster than the traditional computer-aided search method.
Journal ArticleDOI
Analyzing Reaction Rates with the Distortion/Interaction‐Activation Strain Model
TL;DR: This model has been applied to reactions of all types in both organic and inorganic chemistry, including substitutions and eliminations, cycloadditions, and several types of organometallic reactions.
Journal ArticleDOI
Molecular graph convolutions: moving beyond fingerprints
TL;DR: In this article, molecular graph convolutions, a machine learning architecture for learning from undirected graphs, specifically small molecules, are described. But they do not outperform all fingerprint-based methods, and they represent a new paradigm in ligand-based virtual screening with exciting opportunities for future improvement.