Academic Journal
Atomic Uncertainty Pinpoints Critical Failure Structures for Trustworthy Molecular Property Prediction.
| Τίτλος: | Atomic Uncertainty Pinpoints Critical Failure Structures for Trustworthy Molecular Property Prediction. |
|---|---|
| Συγγραφείς: | Qian J; School of Mathematics and Statistics, Central South University, Changsha410083, People's Republic of China., Luo Y; Institute of Interdisciplinary Integrative Medicine Research, Shanghai University of Traditional Chinese Medicine, Shanghai201203, People's Republic of China., Zhou Q; School of Mathematics and Statistics, Central South University, Changsha410083, People's Republic of China., Guan X; Institute of Interdisciplinary Integrative Medicine Research, Shanghai University of Traditional Chinese Medicine, Shanghai201203, People's Republic of China. |
| Πηγή: | Journal of chemical information and modeling [J Chem Inf Model] 2026 Jul 27; Vol. 66 (14), pp. 8076-8086. Date of Electronic Publication: 2026 Jul 14. |
| Τύπος έκδοσης: | Journal Article |
| Γλώσσα: | English |
| Στοιχεία περιοδικού: | Publisher: American Chemical Society Country of Publication: United States NLM ID: 101230060 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1549-960X (Electronic) Linking ISSN: 15499596 NLM ISO Abbreviation: J Chem Inf Model Subsets: MEDLINE |
| Imprint Name(s): | Original Publication: Washington, D.C. : American Chemical Society, c2005- |
| Ιατρικοί όροι (MeSH): | Models, Molecular* , Uncertainty* , Drug Discovery*, Monte Carlo Method ; Solubility |
| Περίληψη: | Accurate prediction of molecular properties is essential for accelerating drug discovery, yet current deep learning methods generally lack reliable uncertainty estimates, particularly at the atomic level. A limitation of current approaches is the reliance on global molecular uncertainty, which frequently masks localized ambiguities by averaging the uncertainty across the entire molecule. To address this, we introduce AUINet, an atomic uncertainty-aware iterative network that quantifies and refines uncertainty at the atomic scale. Built on a D-MPNN architecture, AUINet uses Monte Carlo dropout to estimate atom-level uncertainty and iteratively refines atomic features through uncertainty-guided updates. Comprehensive evaluations show that AUINet outperforms state-of-the-art models on molecular property benchmarks and protein-protein interaction inhibitor tasks under low-data conditions, all without requiring extensive pretraining. Crucially, rejection sampling experiments reveal that atom uncertainty provides a more robust error signal than classic molecular uncertainty. More importantly, AUINet provides chemically interpretable insights by pinpointing specific functional groups and structural motifs that contribute most to prediction uncertainty, as validated in solubility prediction and activity-cliff analysis. Overall, AUINet's precise localization of atomic-level uncertainty establishes a new paradigm for trustworthy molecular property prediction. |
| Entry Date(s): | Date Created: 20260714 Date Completed: 20260805 Latest Revision: 20260805 |
| Update Code: | 20260805 |
| DOI: | 10.1021/acs.jcim.6c00973 |
| PMID: | 42447216 |
| Βάση Δεδομένων: | MEDLINE |
καταχωρήστε σχόλιο πρώτοι!