InChI version 1.06: now more than 99.99% reliable

Abstract

The software for the IUPAC Chemical Identifier, InChI, is extraordinarily reliable. It has been tested on large databases around the world, and has proved itself to be an essential tool in the handling and integration of large chemical databases. InChI version 1.05 was released in January 2017 and version 1.06 in December 2020. In this paper, we report on the current state of the InChI Software, the details of the improvements in the v1.06 release, and the results of a test of the InChI run on PubChem, a database of more than a hundred million molecules. The upgrade introduces significant new features, including support for pseudo-element atoms and an improved description of polymers. We expect that few, if any, applications using the standard InChI will need to change as a result of the changes in version 1.06. Numerical instability was discovered for 0.002% of this database, and a small number of other molecules were discovered for which the algorithm did not run smoothly. On the basis of PubChem data, we can demonstrate that InChI version 1.05 was 99.996% accurate, and InChI version 1.06 represents a step closer to perfection. Finally, we look forward to future releases and extensions for the InChI Chemical identifier.

PeakForest: a multi-platform digital infrastructure for interoperable metabolite spectral data and metadata management

“…The first section presents general compound information (common name, synonyms, molecular formula, accurate and average masses). The second section is related to structural data with common numerical molecule representations in MDL Molfile (Dalby et al, 1992 ), Canonical SMILES (O’Boyle, 2012 ), InChI and InChIKey formats (Goodman et al, 2021 ; Southan, 2013 ) and 2D and 3D molecule images. A “cross-reference identifiers” section includes four selected external chemical compounds bank references, with ChEBI (Hastings et al, 2016 ), PubChem (Kim et al, 2021 ), KEGG (Kanehisa et al, 2016 ), and HMDB (Wishart et al, 2018 ) and a modular system to add any Web hyperlinks from specific biological knowledge banks or metabolic networks databases.…”

SELFIES and the future of molecular string representations

“…However, this behavior might change in future versions. 60 In practice, it has been found that I n C h I performs worse than Smiles in ML-based applications, likely due to the above-mentioned reasons. 54 …”

“…In this section, we discuss the challenges and prospects of extending Selfies beyond organic chemistry. In contrast to organic molecules, 60 transition metal, lanthanide, actinide, and main-group metal compounds are difficult to handle with current digital molecular representations 28 due to special bonding situations and intricate 3D structures, combined with technical limitations that have evolved for historical reasons. Most problems trace back to (1) the assumption that bonding is localized and thus can be described with valence bond (VB) theory, (2) the non-explicit representation of terminal hydrogen atoms, which are added to the heavy (non-H) atoms based on rules derived from VB models in an approach called “implicit hydrogens,” and (3) the inability to describe stereochemistry that goes beyond the usual restrictions of organic chemistry, i.e., stereogenic carbon centers plus some cases of cis / trans isomerism in C=C double bonds and cumulenes.…”

Exploration of Ultralarge Compound Collections for Drug Discovery

“…SciWalker integrates structure-based chemical ontologies with other ontologies resulting in improved associations and classifications of scientific content. Once identified, chemical entities are converted into SMILES and InChIKeys , which render the documents searchable by structure, substructure, or structural similarity.…”