ChemTorch: A Deep Learning Framework for Benchmarking and Developing Chemical Reaction Property Prediction Models

TL;DR

ChemTorch is a new deep learning framework designed to improve the development and benchmarking of models that predict chemical reaction properties.

Contribution

ChemTorch introduces a modular, open-source framework for reproducible benchmarking and model development in chemical reaction property prediction.

Findings

Structurally informed models show clear advantages in barrier-height prediction.

Performance drops significantly under out-of-distribution conditions.

Comparison of four modalities on RDB7 highlights the need for rigorous benchmarking.

Abstract

Modeling of chemical
reactions is essential for understanding kinetic
mechanisms and predicting possible outcomes of reacting systems. Quantum
mechanical calculations are accurate but often prohibitively expensive.
Deep learning has emerged as a faster alternative, but progress is
slowed by a fragmented software ecosystem that hinders reuse, fair
comparison, and reproducibility. We present ChemTorch, an open-source
framework that streamlines model development, experimentation, hyperparameter
tuning, and benchmarking through modular pipelines, standardized configuration,
and built-in data splitters for in- and out-of-distribution evaluation.
We envision ChemTorch as a foundation for community-driven method
development and reproducible benchmarking in chemical reaction modeling.
As a first step toward unified benchmarks, we compare four representative
modalities for barrier-height prediction on the RDB7 data set, including
fingerprint-, sequence-, graph-, and 3D-based approaches. Our results
highlight clear advantages of structurally informed models and sharp
performance drops under out-of-distribution conditions, highlighting
the importance of rigorous benchmarking.

Figures