Interpretation of QSAR Models by Coloring Atoms According to Changes in Predicted Activity: How Robust Is It?
Abstract
Most chemists would agree that the ability to interpret a quantitative structure-activity relationship (QSAR) model is as important as the ability of the model to make accurate predictions. One type of interpretation is coloration of atoms in molecules according to the contribution of each atom to the predicted activity, as in "heat maps". The ability to determine which parts of a molecule increase the activity in question and which decrease it should be useful to chemists who want to modify the molecule. For that type of application, we would hope the coloration to not be particularly sensitive to the details of model building. In this Article, we examine a number of aspects of coloration against 20 combinations of descriptors and QSAR methods. We demonstrate that atom-level coloration is much less robust to descriptor/method combinations than cross-validated predictions. Even in ideal cases where the contribution of individual atoms is known, we cannot always recover the important atoms for some descriptor/method combinations. Thus, model interpretation by atom coloration may not be as simple as it first appeared.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Cites
Breiman, L. (2001) · Machine Learning · doi:10.1023/a:1010933404324 · Cited by 133,538 · see citing works
XGBoost: A Scalable Tree Boosting System
Chen, T., & Guestrin, C. (2016) · Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining · doi:10.1145/2939672.2939785 · Cited by 54,569 · see citing works
Extended-Connectivity Fingerprints
Rogers, D. O. W., & Hahn, M. (2010) · Journal of Chemical Information and Modeling · doi:10.1021/ci100050t · Cited by 7,597 · see citing works
LIBLINEAR: A Library for Large Linear Classification
Fan, R.-E., Chang, K., Hsieh, C., Xiang-rui, W., & Lin, C. (2008) · doi:10.5555/1390681.1442794 · Cited by 6,680 · see citing works
ChEMBL: a large-scale bioactivity database for drug discovery
Gaulton, A., Bellis, L. J., Bento, A. P., Chambers, J., Davies, M., Hersey, A., Light, Y., McGlinchey, S., Michalovich, D., Al-Lazikani, B., et al. (2012) · Nucleic Acids Research · doi:10.1093/nar/gkr777 · Cited by 4,382 · see citing works
Random Forest: A Classification and Regression Tool for Compound Classification and QSAR Modeling
Svetnik, V., Liaw, A. I., Tong, C., Culberson, J. C., Sheridan, R. P., & Feuston, B. P. (2003) · Journal of Chemical Information and Computer Sciences · doi:10.1021/ci034160g · Cited by 3,613 · see citing works
Self-citationThe pls Package: Principal Component and Partial Least Squares Regression in R
Mevik, B., & Wehrens, R. (2007) · Journal of Statistical Software · doi:10.18637/jss.v018.i02 · Cited by 1,826 · see citing works
Deep Neural Nets as a Method for Quantitative Structure–Activity Relationships
Ma, J., Sheridan, R. P., Liaw, A., Dahl, G. E., & Svetnik, V. (2015) · Journal of Chemical Information and Modeling · doi:10.1021/ci500747n · Cited by 1,159 · see citing works
Self-citationAtom pairs as molecular features in structure-activity studies: definition and applications
Carhart, R. E., Smith, D. H., & Venkataraghavan, R. (1985) · Journal of Chemical Information and Computer Sciences · doi:10.1021/ci00046a002 · Cited by 1,120 · see citing works
Deep Learning in Drug Discovery
Gawehn, E., Hiss, J. A., & Schneider, G. (2016) · Molecular Informatics · doi:10.1002/minf.201501008 · Cited by 723 · see citing works
Extreme Gradient Boosting as a Method for Quantitative Structure–Activity Relationships
Sheridan, R. P., Wang, W. M., Liaw, A., Ma, J., & Gifford, E. M. (2016) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.6b00591 · Cited by 544 · see citing works
Self-citationTime-Split Cross-Validation as a Method for Estimating the Goodness of Prospective Prediction.
Sheridan, R. P. (2013) · Journal of Chemical Information and Modeling · doi:10.1021/ci400084k · Cited by 376 · see citing works
Self-citationNilakantan, R., Bauman, N., Dixon, J. S., & Venkataraghavan, R. (1987) · Journal of Chemical Information and Computer Sciences · doi:10.1021/ci00054a008 · Cited by 367 · see citing works
Chemical Similarity Using Physiochemical Property Descriptors
Kearsley, S. K., Sallamack, S. J., Fluder, E. M., Andose, J. D., Mosley, R. T., & Sheridan, R. P. (1996) · Journal of Chemical Information and Computer Sciences · doi:10.1021/ci950274j · Cited by 267 · see citing works
Self-citationInterpretation of Quantitative Structure-Activity Relationship Models: Past, Present, and Future
Polishchuk, P. (2017) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.7b00274 · Cited by 236 · see citing works
Similarity maps - a visualization strategy for molecular fingerprints and machine-learning methods
Riniker, S., & Landrum, G. A. (2013) · Journal of Cheminformatics · doi:10.1186/1758-2946-5-43 · Cited by 228 · see citing works
Boosting: An Ensemble Learning Tool for Compound Classification and QSAR Modeling
Svetnik, V., Wang, T., Tong, C., Liaw, A. I., Sheridan, R. P., & Song, Q. (2005) · Journal of Chemical Information and Modeling · doi:10.1021/ci0500379 · Cited by 219 · see citing works
Self-citationMarchese Robinson, R. L., Palczewska, A., Palczewski, J., & Kidley, N. (2017) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.6b00753 · Cited by 167 · see citing works
Universal Approach for Structural Interpretation of QSAR/QSPR Models
Polishchuk, P. G., Kuz'min, V. E., Artemenko, A. G., & Muratov, E. N. (2013) · Molecular Informatics · doi:10.1002/minf.201300029 · Cited by 76 · see citing works
Kernel-Based Partial Least Squares: Application to Fingerprint-Based QSAR with Model Visualization
An, Y., Sherman, W., & Dixon, S. L. (2013) · Journal of Chemical Information and Modeling · doi:10.1021/ci400250c · Cited by 68 · see citing works
Interpreting linear support vector machine models with heat map molecule coloring
Rosenbaum, L., Hinselmann, G., Jahn, A., & Zell, A. (2011) · Journal of Cheminformatics · doi:10.1186/1758-2946-3-11 · Cited by 57 · see citing works
Interpretability of SAR/QSAR Models of any Complexity by Atomic Contributions
Marcou, G., Horvath, D., Solov’ev, V. P., Arrault, A., Vayer, P., & Varnek, A. A. (2012) · Molecular Informatics · doi:10.1002/minf.201100136 · Cited by 45 · see citing works
Benchmarking the Predictive Power of Ligand Efficiency Indices in QSAR
Cortés‐Ciriano, I. (2016) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.6b00136 · Cited by 42 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
Drug discovery with explainable artificial intelligence
Jiménez-Luna, J., Grisoni, F., & Schneider, G. (2020) · Nature Machine Intelligence · doi:10.1038/s42256-020-00236-4 · Cited by 1,063 · see citing works
Muratov, E. N., Bajorath, J., Sheridan, R. P., Tetko, I. V., Filimonov, D., Poroikov, V., Oprea, T. I., Baskin, I. I., Varnek, A., Roitberg, A., et al. (2020) · Chemical Society Reviews · doi:10.1039/d0cs00098a · Cited by 913 · see citing works
Self-citationArtificial intelligence in chemistry and drug design
Brown, N., Ertl, P., Lewis, R., Luksch, T., Reker, D., & Schneider, N. (2020) · Journal of Computer-Aided Molecular Design · doi:10.1007/s10822-020-00317-x · Cited by 165 · see citing works
Model agnostic generation of counterfactual explanations for molecules
Wellawatte, G. P., Seshadri, A., & White, A. D. (2022) · Chemical Science · doi:10.1039/d1sc05259d · Cited by 143 · see citing works
Coloring Molecules with Explainable Artificial Intelligence for Preclinical Relevance Assessment
Jiménez-Luna, J., Skalic, M., Weskamp, N., & Schneider, G. (2021) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.0c01344 · Cited by 96 · see citing works
- Authors
- Sheridan, R. P.
- Affiliations
- Merck & Co., Inc., Rahway, NJ, USA (United States) (US)
- Corresponding author
- Sheridan, Robert P.
- ORCID
- Sheridan, Robert P.
- Venue
- Journal of Chemical Information and Modeling
- Publisher
- American Chemical Society (ACS)
- Published
- 2019-02-19
- Volume
- 59
- Issue
- 4
- Pages
- 1324-1337
- ISSN
- 1549-9596 (ISSN-L)
- 1549-960X
- OpenAlex
- W2916327454
- Cited by
- 84
- References
- 23
- Citations by year
- 2026: 12 · 2025: 8 · 2024: 12 · 2023: 17 · 2022: 16 · 2021: 11 · 2020: 8
- FWCI
- 6.11
- Type
- journal article
- Language
- en
- Open access
- closed
- Licence
- Locations
- Journal of Chemical Information and Modeling (publishedVersion)
- PubMed (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Computational Theory and Mathematics › Computational Drug Discovery Methods
- Physical Sciences › Physics and Astronomy › Atomic and Molecular Physics, and Optics › Advanced Chemical Physics Studies
- Physical Sciences › Materials Science › Materials Chemistry › Machine Learning in Materials Science
- Keywords
- heatmaps, robustness, cross-validation, molecular descriptors, structure-activity relationship
- MeSH terms
- Machine Learning, Computer Simulation, Humans, Quantitative Structure-Activity Relationship, Workflow
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
