A possible extension to the RInChI as a means of providing machine readable process data

Abstract

The algorithmic, large-scale use and analysis of reaction databases such as Reaxys is currently hindered by the absence of widely adopted standards for publishing reaction data in machine readable formats. Crucial data such as yields of all products or stoichiometry are frequently not explicitly stated in the published papers and, hence, not reported in the database entry for those reactions, limiting their usefulness for algorithmic analysis. This paper presents a possible extension to the IUPAC RInChI standard via an auxiliary layer, termed ProcAuxInfo, which is a standardised, extensible form in which to report certain key reaction parameters such as declaration of all products and reactants as well as auxiliaries known in the reaction, reaction stoichiometry, amounts of substances used, conversion, yield and operating conditions. The standard is demonstrated via creation of the RInChI including the ProcAuxInfo layer based on three published reactions and demonstrates accurate data recoverability via reverse translation of the created strings. Implementation of this or another method of reporting process data by the publishing community would ensure that databases, such as Reaxys, would be able to abstract crucial data for big data analysis of their contents.

Molecular representations in AI-driven drug discovery: a review and practical guide

“…A proposed further extension to RInChI, ProcAuxInfo, enables the storage of metadata relating to yields, temperature, concentration, and other reaction conditions [71]. RInChI offers an alternative to Reaction SMILES that enables the identification of duplicate reactions, as the order in which molecules are listed in Reaction SMILES is arbitrary.…”

“…RInChI offers an alternative to Reaction SMILES that enables the identification of duplicate reactions, as the order in which molecules are listed in Reaction SMILES is arbitrary. Hashing the RInChI to yield the RInChI key provides a powerful tool for efficiently indexing and searching reaction data [18,71]. However, there is no SMARTS or SMIRKS equivalent for RInChI, limiting its use in substructure searching and in encoding generic chemical transformations.…”

Reaction SPL – extension of a public document markup standard to chemical reactions

“…In contrast, InChI [ 11 ], for example, is a very specific model, thus it is more difficult to broaden it to, e.g., general reaction descriptions such as inclusion of reaction conditions, yields etc. The reaction identifier RInChI (Reaction InChI) [ 12 ], which is an application of InChI since it is a concatenation of individual InChI strings of reactants and products, offers the possibility to add additional information in an Auxiliary Information section, currently limited to taking the AuxInfo of InChIs in order to enable reconstruction of the full molecule (a proposed expansion of the RInChI AuxInfo to allow providing machine readable process data [ 13 ] does not seem to have been implemented in the RInChI tool downloadable from the InChI Trust [ 14 ]). Still, RInChI is constitutively limited by the underlying model of InChI.…”

Chemical data intelligence for sustainable chemistry

“…75,[80][81][82] Similar structures of semantic web will be essential to enable sustainability assessment within reaction networks, where temperatures, yields, solvents, and reaction stoichiometry should be recorded and linked with each reaction. 83 An entity-based data format would allow us to link further information (e.g. data on waste streams and their compositions, or on regional availabilities of renewable energy) in a modular way, paving the way for more holistic considerations in future reaction planning.…”