Tools
Generate 2D images from SMILES.
Convert between molecule file formats.
Generate product libraries.
Count functional groups (sub-structures).
Screen molecules by functional group count.
Fragment molecules for mass spec analysis.
Search ChemDB by monoisotopic mass and substructure filtering.
Predict activities of small molecules against a large set of protein targets
Datasets
Datasets for training and testing machine learning and other algorithms.
Dataset of chemical reactivities
Publications
Relevant scientific articles published by our team.
Miller R.J., Dashuta A.E., Rudisill B., Van Vranken D., Baldi P.
Mechanism-Aware Deep Learning for Polar Reaction Prediction Journal of
American Chemical Society, 2025, Oct 2025
Tavakoli M., Miller R.J., Angel M.C., Pfeiffer M.A., Gutman E.S., Mood A.D., Van Vranken D., Baldi P.
PMechDB: A Public Database of Elementary Polar Reaction Steps Journal of
Chemical Information and Modeling, 64 (4), Mar 2024
Tavakoli M., Baldi P., Carlton A.M., Chiu Y.T., Shmakov A., Van Vranken D.
AI for Interpretable Chemistry: Predicting Radical Mechanistic Pathways via
Contrastive Learning NeurIPS Proceedings 36 (NeurIPS 2023), 2023
Tavakoli M., Chiu Y.T.,Baldi P.,Carlton A.M., Van Vranken D., RMechDB: A Public Database of Elementary Radical Reaction Steps,
Journal of Chemical Information and Modeling, 63 (4), Feb 2023
If you use any data or tools from the ChemDB Portal, please cite the following article:
Chen, J. H., Linstead, E., Swamidass, S. J., Wang, D. & Baldi, P. ChemDB update-full-text search and virtual chemical space. Bioinformatics 23, 2348-2351 (2007).
Molecules
Find a molecule by its name, structure, or similarity to another molecule and filter the results.
Interactively deconstruct a target molecule into possible chemical precursors and reassemble them into a combinatorial library of real or virtual molecules around the target.
Calculate or predict molecular properties other than 3D structure.
Predict aqueous solubility of small molecules using
deep learning and ensembles.
Predict 3D molecular structures using open
crystallography libraries.
Predict 3D molecular structures.
Inner- and Outer Recursive Neural Networks for Chemoinformatics
Reactions
Polar Mechanistic Reaction Predictor
Radical Mechanistic Reaction Predictor
Polar Mechanistic Reaction Database:
PMechDB is a live platform for aggregating, curating, and distributing chemical reactions in the form of polar elementary steps to accelerate research in chemoinformatics and polar reaction modeling. The PMechDB platform is designed to facilitate training deep learning and other AI models in data-driven workflows using its tabular data, with no need for additional pre-processing steps. It provides a unified model that ought to facilitate data sharing, model building, dissemination, and publications. We encourage the community to explore and use the PMechDB data and functionalities, and contribute to its expansion.
PMechDB is a live platform for aggregating, curating, and distributing chemical reactions in the form of polar elementary steps to accelerate research in chemoinformatics and polar reaction modeling. The PMechDB platform is designed to facilitate training deep learning and other AI models in data-driven workflows using its tabular data, with no need for additional pre-processing steps. It provides a unified model that ought to facilitate data sharing, model building, dissemination, and publications. We encourage the community to explore and use the PMechDB data and functionalities, and contribute to its expansion.
Radical Mechanistic Reaction Predictor
Radical Mechanistic Reaction Database:
RMechDB is a live platform for aggregating, curating, and distributing chemical reactions in the form of elementary radical steps to accelerate research in chemoinformatics and radical reaction modeling.
RMechDB is a live platform for aggregating, curating, and distributing chemical reactions in the form of elementary radical steps to accelerate research in chemoinformatics and radical reaction modeling.
Predict reaction outcomes and mechanisms using machine learning.
Learn and practice reactions, syntheses, and mechanisms interactively with support for: automated generation of problems, curved-arrow mechanism diagrams, and inquiry-based learning.
Previous system to predict reaction outcomes and
mechanisms using machine learning.
Predict the mapping of reactant atoms to product atoms for chemical reactions.
Download & Documentation
Download
Download entire set of chemical isomers contained within ChemDB.
Implementation
System implementation materials such as the database schema with data definition and source / vendor information table.
Analysis
Data analysis tables and charts based upon ChemDB contents.
Acknowledgements
JME Editor (Peter Ertl, Novartis)
All of the vendors who have supplied their chemical information catalogs that comprise the core data beneath ChemDB.
The source information table includes a complete listing.

University of California, Irvine
Institute for Genomics and Bioinformatics