ISAAC: Prior-aligned structural sensitivity auditing for drug–target interaction models
Artificial Intelligence in the Life Sciences · 2026-12 · 10 · 100180
DOI: 10.1016/j.ailsci.2026.100180 ↗Kurz gefasst
Fachartikel in der Zeitschrift Artificial Intelligence in the Life Sciences, 2026-12. Verfasst von Barbara Tarantino und 3 weiteren Autor:innen. Die Arbeit nennt 42 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Deep learning models for drug–target interaction (DTI) prediction often achieve strong benchmark performance while relying on input patterns that are not captured by standard accuracy-based evaluation. We introduce ISAAC (Intervention-based Structural Auditing ApproaCh for Model Evaluation), a post-hoc framework that evaluates prior-relative structural sensitivity by probing frozen models through matched prior-aligned and prior-misaligned input-level interventions, independently of predictive accuracy. Applied to three sequence-based DTI architectures on the Davis kinase benchmark, ISAAC reveals statistically significant differences in prior-alignment ratios across models operating within a comparable predictive regime, differences that are not explained by predictive performance alone and that are robust to the choice of perturbation operator. These discrepancies are not reflected by the predictive ranking and motivate the use of post-hoc structural auditing as a complement to standard performance evaluation in DTI modeling and, more broadly, in scientific machine learning settings where structured priors are available.
Zitieren
APA
Tarantino, B., Kim, S., Lu, Y., & Giudici, P. (2026). ISAAC: Prior-aligned structural sensitivity auditing for drug–target interaction models. Artificial Intelligence in the Life Sciences, 10, 100180. https://doi.org/10.1016/j.ailsci.2026.100180
Vancouver
Tarantino B, Kim S, Lu Y, Giudici P. ISAAC: Prior-aligned structural sensitivity auditing for drug–target interaction models. Artificial Intelligence in the Life Sciences. 2026;10:100180. doi:10.1016/j.ailsci.2026.100180
BibTeX
@article{Tarantino2026,
author = {Tarantino, Barbara and Kim, Sun and Lu, Yijingxiu and Giudici, Paolo},
title = {ISAAC: Prior-aligned structural sensitivity auditing for drug–target interaction models},
journal = {Artificial Intelligence in the Life Sciences},
year = {2026},
volume = {10},
pages = {100180},
doi = {10.1016/j.ailsci.2026.100180}
}
EUVIMED