Comment on the paper “indexing molecules with chemical graph Identifiers” by Elisabeth Gregori‐Puigjané, Rut Garriga‐Sust, and Jordi Mestres

The authors of the article ‘‘Indexing Molecules with Chemical Graph Identifiers’’ compare their new indexing algorithm with two other established structure key generators—NIST InChI keys and the structure hashcodes of the Cactvs toolkit. In the publication, a total of three examples are given where the authors claim that the Cactvs software fails, or lacks specific features, and thus their solution yields superior results in these cases. All three examples are demonstrably incorrect. This can be easily verified by anybody interested with the original test data and the free academic version of the Cactvs toolkit. The claims of a hashcode collision and the inability to compute the same hashcode for an equivalent structure are potentially extremely damaging, as Cactvs hashcodes are used as primary structure keys in important databases. This letter was written to refute these claims. The three listed problem cases and shortcomings are: