ChemAudit: An Open-Source Chemical Structure Validation Suite  

Data quality in chemistry remains one of the biggest bottlenecks in cheminformatics, drug discovery, and machine learning for chemistry. Issues such as incorrect structural representations, undefined stereocenters, PAINS-flagged1 compounds, and inconsistent standardisation can quietly undermine the reliability of downstream models and analyses.

ChemAudit was built to address this. It’s a free, open-source web platform that brings structure validation, standardisation, structural alert screening, and quality scoring together in one clear, user-friendly interface. No command-line experience needed.

What it does: 

  • Runs 15+ validation checks covering parsability, valence, stereochemistry, and representation consistency
  • Screens against 480+ PAINS patterns and 700+ pharmaceutical alert filters sourced from BMS, Glaxo, Dundee, and other ChEMBL2 collections
  • Scores ML-readiness (0–100) by testing 451 molecular descriptors and 7 fingerprint types
  • Evaluates drug-likeness via Lipinski3, QED4, Veber5, Ghose6, and Muegge 7 rules
  • Predicts ADMET properties, including synthetic accessibility, solubility, and CNS penetration
  • Standardises structures using the ChEMBL pipeline8 (salt stripping, tautomer canonicalization, charge normalisation)
  • Assesses natural product likeness with scaffold analysis

Built for scale: Batch processing supports up to 1M molecules with real-time WebSocket progress tracking. Results can be exported to CSV, Excel, SDF, JSON, and PDF.

Built on proven tools: RDKit9, MolVS10, and the ChEMBL structure pipeline power the backend. React and RDKit.js deliver interactive 2D depictions with atom-level issue highlighting on the frontend.

ChemAudit is designed for database curators, ML researchers, medicinal chemists, and natural products scientists who need reliable, standardised chemical data without the overhead of stitching together disparate CLI tools or licensing commercial software.

Self-hosted and MIT-licensed. Try it, break it, extend it.

Available at:

1. Baell, J. B. & Holloway, G. A. New substructure filters for removal of pan assay interference compounds (PAINS) from screening libraries and for their exclusion in bioassays. J. Med. Chem. 53, 2719–2740 (2010).

2. Gaulton, A. et al. The ChEMBL database in 2017. Nucleic Acids Res. 45, D945–D954 (2017).

3. Lipinski, C. A., Lombardo, F., Dominy, B. W. & Feeney, P. J. Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv. Drug Deliv. Rev. 23, 3–25 (1997).

4. Bickerton, G. R., Paolini, G. V., Besnard, J., Muresan, S. & Hopkins, A. L. Quantifying the chemical beauty of drugs. Nat. Chem. 4, 90–98 (2012).

5. Veber, D. F. et al. Molecular properties that influence the oral bioavailability of drug candidates. J. Med. Chem. 45, 2615–2623 (2002).

6. Ghose, A. K., Viswanadhan, V. N. & Wendoloski, J. J. A knowledge-based approach in designing combinatorial or medicinal chemistry libraries for drug discovery. 1. A qualitative and quantitative characterization of known drug databases. J. Comb. Chem. 1, 55–68 (1999).

7. Muegge, I., Heald, S. L. & Brittelli, D. Simple selection criteria for drug-like chemical matter. J. Med. Chem. 44, 1841–1846 (2001).

8. Bento, A. P. et al. An open source chemical structure curation pipeline using RDKit. J. Cheminform. 12, 51 (2020).

9. Landrum, G. RDKit: Open-source cheminformatics. (2006).

10. Swain, M. MolVS: Molecule Validation and Standardization. (Github).


Posted on by Theo Bender