Extraction of organic chemistry grammar from unsupervised learning of chemical reactions
277 Citations•2021•
Philippe Schwaller, Benjamin Hoover, Jean‐Louis Reymond
This work demonstrates that Transformer Neural Networks learn atom-mapping information between products and reactants without supervision or human labeling, and provides the missing link between data-driven and rule-based approaches for numerous chemical reaction tasks.
Abstract
<jats:p>RXNmapper constructs coherent atom-mapping rules from raw chemical reactions using unsupervised training of neural networks.</jats:p>