pipette
ENEnglish

Improving molecular property prediction with score-based diffusion models

Julien Horwood, Dino Oglić, José Miguel Hernández-Lobato

Revista con revisión por pares

En palabras de los autores

Abstract We explore diffusion-based training as a flexible approach to molecular property prediction in settings shaped by chemical, biophysical, and latent heterogeneity. To address this, we build on score-based generative diffusion models and introduce a conditional diffusion module for predictive tasks, equipped with a joint training and sampling guidance scheme and an optional gated residual mechanism that regulates feature- vs target-driven information flow across noise levels. The module acts as a plug-in learning block for existing molecular encoders, making it complementary to advances in representation learning. Across diverse feature extractors and graph neural networks, diffusion-based training improves over conventional maximum-likelihood objectives on small-molecule regression benchmarks and matches or exceeds strong baselines on multitask regression and classification datasets. It also produces well-calibrated predictive uncertainty as measured by prediction interval coverage, enables confidence-guided ranking, and performs well in challenging settings including activity cliffs and class-imbalanced multitask prediction. Overall, these results position diffusion as a promising representation-agnostic approach for effective molecular property prediction.

Resultado principalEl resumen no menciona limitaciones.

Apareció: miércoles, 18 de noviembre. Nature Communications. Revista con revisión por pares.

DOI: 10.1038/s41467-026-79240-y