ArticleProceedings of the National Academy of Sciences of the United States of America2019
Using attribution to decode binding mechanism in neural network models for chemistry.
Article in Proceedings of the National Academy of Sciences of the United States of America, 2019. The graph could read no effect estimate from its abstract, so it casts no vote on the map. Cited by 34 papers.
What it found
Each row is one number read from the abstract, on the scale the paper reported it, with its interval. Left of the dashed line favours the treatment, right favours the comparator. Under each row is the sentence it came from. New to these charts? A ten-minute tutorial.
The abstract states no effect estimate the extractor could read, or names no intervention and outcome on the map, so this paper lights no cell and moves no belief. It is still indexed, cited and linked below.
The trial behind it
Trials whose registry record cites this paper, or whose number appears in the abstract. A trial that started after this paper was published is citing it as background, not reporting it.
Neither the registry nor the abstract names a trial number. If this is a trial report, that itself is worth knowing.
Who cites it
34 citing papers in PubMed.
- An interpretable framework applying protein words to predict protein-small molecule complementary pairing rules.Chemical science · 2026Article
- Deep-learning-based virtual screening of antibacterial compounds.Nature biotechnology · 2026Article
- Systematic validation of graph neural network explanations against adverse outcome pathway reactive centers for skin sensitization.Journal of cheminformatics · 2026Article
- Mol2Raman: a graph neural network model for predicting Raman spectra from SMILES representations.Digital discovery · 2026Article
- Structure-Aware Compound-Protein Affinity Prediction via Graph Neural Networks with Group Lasso Regularization.Computational and structural biotechnology journal · 2026Article
- Peering inside the black box by learning the relevance of many-body functions in neural network potentials.Nature communications · 2025Article
- Bridging traditional and contemporary approaches in computational medicinal chemistry: opportunities for innovation in drug discovery.RSC medicinal chemistry · 2025Review
- ACES-GNN: can graph neural network learn to explain activity cliffs?Digital discovery · 2025Article
- PerturbNet predicts single-cell responses to unseen chemical and genetic perturbations.Molecular systems biology · 2025Article
- Analyzing Atomic Interactions in Molecules as Learned by Neural Networks.Journal of chemical theory and computation · 2025Article
- Interpretable Deep-Learning pJournal of chemical information and modeling · 2025Article
- Advances, opportunities, and challenges in methods for interrogating the structure activity relationships of natural products.Natural product reports · 2024Review
- Interpreting Neural Network Models for Toxicity Prediction by Extracting Learned Chemical Features.Journal of chemical information and modeling · 2024Article
- Enhancing property and activity prediction and interpretation using multiple molecular graph representations with MMGX.Communications chemistry · 2024Article
- Machine learning coupled with causal inference to identify COVID-19 related chemicals that pose a high concern to drinking water.iScience · 2024Article
- Impossibility theorems for feature attribution.Proceedings of the National Academy of Sciences of the United States of America · 2024Article
- Integrating Explainability into Graph Neural Network Models for the Prediction of X-ray Absorption Spectra.Journal of the American Chemical Society · 2023Article
- Automatic identification of chemical moieties.Physical chemistry chemical physics : PCCP · 2023Article
- Exploring the ability of machine learning-based virtual screening models to identify the functional groups responsible for binding.Journal of cheminformatics · 2023Article
- Explaining compound activity predictions with a substructure-aware loss for graph neural networks.Journal of cheminformatics · 2023Article
Corrections and comments
PubMed lists nothing against this paper. Absence here is not a guarantee, only a check that was made.
Authors and funding
5 authors.
Funding
No grant is acknowledged in the PubMed record.
Abstract
Deep neural networks have achieved state-of-the-art accuracy at classifying molecules with respect to whether they bind to specific protein targets. A key breakthrough would occur if these models could reveal the fragment pharmacophores that are causally involved in binding. Extracting chemical details of binding from the networks could enable scientific discoveries about the mechanisms of drug actions. However, doing so requires shining light into the black box that is the trained neural network model, a task that has proved difficult across many domains. Here we show how the binding mechanism learned by deep neural network models can be interrogated, using a recently described attribution method. We first work with carefully constructed synthetic datasets, in which the molecular features responsible for "binding" are fully known. We find that networks that achieve perfect accuracy on held-out test datasets still learn spurious correlations, and we are able to exploit this nonrobustness to construct adversarial examples that fool the model. This makes these models unreliable for accurately revealing information about the mechanisms of protein-ligand binding. In light of our findings, we prescribe a test that checks whether a hypothesized mechanism can be learned. If the test fails, it indicates that the model must be simplified or regularized and/or that the training dataset requires augmentation.
Indexed as
Identifiers
What OpenQuestion holds
Registered trials
Read under generation 80e0d062 · epoch 390. Bibliography from PubMed, PubMed Central and OpenAlex; grants from NIH RePORTER; trial links from ClinicalTrials.gov; estimates, votes and beliefs from the OpenQuestion graph.