Issue 9, 2024

What can attribution methods show us about chemical language models?

Abstract

Language models trained on molecular string representations have shown strong performance in predictive and generative tasks. However, practical applications require not only making accurate predictions, but also explainability – the ability to explain the reasons and rationale behind the predictions. In this work, we explore explainability for a chemical language model by adapting a transformer-specific and a model-agnostic input attribution technique. We fine-tune a pretrained model to predict aqueous solubility, compare training and architecture variants, and evaluate visualizations of attributed relevance. The model-agnostic SHAP technique provides sensible attributions, highlighting the positive influence of individual electronegative atoms, but does not explain the model in terms of functional groups or explain how the model represents molecular strings internally to make predictions. In contrast, the adapted transformer-specific explainability technique produces sparse attributions, which cannot be directly attributed to functional groups relevant to solubility. Instead, the attributions are more characteristic of how the model maps molecular strings to its latent space, which seems to represent features relevant to molecular similarity rather than functional groups. These findings provide insight into the representations underpinning chemical language models, which we propose may be leveraged for the design of informative chemical spaces for training more accurate, advanced and explainable models.

Graphical abstract: What can attribution methods show us about chemical language models?

Supplementary files

Transparent peer review

To support increased transparency, we offer authors the option to publish the peer review history alongside their article.

View this article’s peer review history

Article information

Article type
Paper
Submitted
25 Mar 2024
Accepted
27 Jun 2024
First published
18 Jul 2024
This article is Open Access
Creative Commons BY license

Digital Discovery, 2024,3, 1738-1748

What can attribution methods show us about chemical language models?

S. Hödl, T. Kachman, Y. Bachrach, W. T. S. Huck and W. E. Robinson, Digital Discovery, 2024, 3, 1738 DOI: 10.1039/D4DD00084F

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. You can use material from this article in other publications without requesting further permissions from the RSC, provided that the correct acknowledgement is given.

Read more about how to correctly acknowledge RSC content.

Social activity

Spotlight

Advertisements