Data-Driven Chemical Reaction Classification with Attention-Based Neural Networks

27 September 2019, Version 1
This content is a preprint and has not undergone peer review at the time of posting.

Abstract

Organic reactions are usually clustered in classes that collect entities undergoing similar structural rearrangement. The classification process is a tedious task, requiring first an accurate mapping of the rearrangement (atom mapping) followed by the identification of the corresponding reaction class template. In this work, we present a transformer-based model that infers reaction classes from the SMILES representation of chemical reactions. The model reaches an accuracy of 93.8 % for a multi-class classification task involving several hundred different classes. The attention weights provided by the model give an insight into what parts of the SMILES strings are taken into account for classification, based solely on data. We study the incorrect predictions of our model and show that it uncovers different biases and mistakes in the underlying data set.

Keywords

machine learning
deep learning
transformer
organic chemistry
organic synthesis
SMILES-Encoded Molecular Structures
SMILES
SMILES string representation
Chemical Reactions
classification

Comments

Comments are not moderated before they are posted, but they can be removed by the site moderators if they are found to be in contravention of our Commenting Policy [opens in a new tab] - please read this policy before you post. Comments should be used for scholarly discussion of the content in question. You can find more information about how to use the commenting feature here [opens in a new tab] .
This site is protected by reCAPTCHA and the Google Privacy Policy [opens in a new tab] and Terms of Service [opens in a new tab] apply.