Making the InChI FAIR and sustainable while moving to inorganics

Area of Science:

• Chemical informatics
• Computational chemistry
• Data standardization

Background:

• The International Chemical Identifier (InChI) is crucial for chemical information exchange and internet-scale structure searching.
• The InChI algorithm's internal workings were previously undocumented, making it a "black box."
• Existing InChI representations are inadequate for molecular inorganic compounds, particularly concerning metal-ligand bonds.

Purpose of the Study:

• To document and improve the InChI algorithm's codebase.
• To enhance the representation of molecular inorganic compounds within the InChI standard.
• To increase transparency and facilitate community contributions to InChI development.

Main Methods:

• Analysis and documentation of the InChI codebase for the v1.07 release.
• Fixing over 3000 bugs and security issues, including nearly 60 from Google OSS-Fuzz.
• Development of new routines for representing inorganic compounds, focusing on metal-ligand bonds.

Main Results:

• Significant improvements in code stability and security through extensive bug fixing.
• Enhanced transparency via code analysis, documentation, and migration to GitHub.
• Proposed novel methods to enable meaningful InChI representation for molecular inorganic compounds.

Conclusions:

• The modernized InChI codebase is more robust, secure, and transparent.
• The proposed routines offer a solution for the long-standing challenge of representing inorganic compounds.
• These advancements will improve the utility of InChI for inorganic chemistry research and data management.