August 2, 2025Open Access

DeepADR: Multi-modal Prediction of Adverse Drug Reaction Frequency by Integrating Early-Stage Drug Discovery Information via Kolmogorov-Arnold Networks

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Deep learning improves adverse drug reaction prediction in drug discovery, suggesting a new approach using chemical and biological data.

DeepADR uses a multi-modal deep learning approach to predict the occurrence and frequency of adverse drug reactions.
The model integrates chemical structures, biological target profiles, and semantic representations from a large language model.
Utilizing a Kolmogorov-Arnold Network allows DeepADR to model complex interactions and improve predictions for new chemical entities.
This framework enhances early-stage drug safety assessments and prioritizes safer candidates, advancing public health.

Wan et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark