Openclatura: An Open and Inspectable IUPAC Naming Tool for Molecular Discovery Workflows
Adrian Mirza ⋅ Kevin Maik Jablonka ⋅ Rostislav Fedorov
Abstract
We introduce Openclatura, an open-source, deterministic tool for generating systematic chemical names from molecular structures. Its rule-based engine records an inspectable execution trace that binds generated name fragments and nomenclature decisions to atoms and bonds in the molecular graph. Candidate rules were developed with executable chemical feedback: generated names are parsed back into structures with OPSIN and checked for round-trip consistency. Across QM9, PubChem, and ZINC22, Openclatura produces structurally valid names for $100\%$, $99.0 \pm 0.0\%$, and $96.5 \pm 0.1\%$ of molecules, respectively, outperforming the neural STOUT baseline while running on a single CPU core. Robomoleculographer exposes the same structured traces as graph-grounded natural-language descriptions. Together, these components provide standardized textual metadata, inspectable naming traces, and a callable representation layer for AI-enabled molecular and materials discovery workflows.
Chat is not available.
Successful Page Load