Deep learning models for drug--target interaction (DTI) prediction often achieve strong benchmark performance without necessarily relying on mechanistically meaningful molecular features, a limitation that standard accuracy-based evaluation cannot detect. We introduce ISAAC (Intervention-based Structural Auditing Approach for Causal Reasoning), a post-hoc framework that evaluates prior-relative structural sensitivity by probing frozen models through matched mechanistic and spurious input-level interventions, independently of predictive accuracy. Applied to three sequence-based DTI architectures on the Davis benchmark, ISAAC reveals approximately 25\% relative differences in reasoning scores across models with comparable AUROC (within around 3\%), stable across training and intervention seeds and two distinct perturbation operators. These discrepancies, undetectable under conventional accuracy metrics, motivate the use of post-hoc structural auditing as a complement to standard performance evaluation in scientific machine learning for molecular modeling.
@article{arxiv.2605.02962,
title = {ISAAC: Auditing Causal Reasoning in Deep Models for Drug-Target Interaction},
author = {Barbara Tarantino and Sun Kim and Yijingxiu Lu and Paolo Giudici},
journal= {arXiv preprint arXiv:2605.02962},
year = {2026}
}