Python Dermal Sensitisation Threshold (dst) High Potency Category Chemical classifier

A python implementation of the "Principles for identification of High Potency Category Chemicals for which the Dermal Sensitisation Threshold (DST) approach should not be applied".

Implementation relies on the structural alerts developed by Enoch et al (2011) (https://doi.org/10.3109/10408444.2011.598141) that were integrated into the OECD QSAR Toolbox. The subset of alerts in addition to other refinements were made to capture the relevant set of alerts that best aligned with the rules described in Roberts et al (2015) (http://dx.doi.org/10.1016/j.yrtph.2015.03.001).

Features

• Rule set implemented

• RDKit-based: Uses RDKit (www.rdkit.org) for molecular analysis and SMARTS pattern matching

Performance

Validated against the set of chemicals listed in Roberts et al (2015)

Installation

This package requires RDKit. Dependencies are captured in the pyproject.toml file.

uv add dst_hpc

Licensing

This implementation is licensed under MIT.

References

• Enoch et al. (2011): A review of the electrophilic reaction chemistry involved in covalent protein binding relevant to toxicity. Crit Rev Toxicol, 41(9):783-802. doi: 10.3109/10408444.2011.598141.
• Roberts et al. (2015): Principles for identification of High Potency Category Chemicals for which the Dermal Sensitisation Threshold (DST) approach should not be applied. Regul Toxicol Pharmacol, 81, 72(3):683-93. doi: 10.1016/j.yrtph.2015.03.001.
• RDKit: https://www.rdkit.org/

Acknowledgements

A handful of SMARTS patterns and ring detection logic were developed with assistance from Claude Haiku (Anthropic), with subsequent code review to verify correctness.