ReactionCode: format for reaction searching, analysis, classification, transform, and encoding/decoding.

Area of Science:

• Chemical informatics
• Computational chemistry

Background:

• Numerous molecular and reaction formats exist, primarily for data exchange and analysis.
• Existing reaction formats are often proprietary, hindering open-source development and accessibility.

Purpose of the Study:

• Introduce ReactionCode, a novel open-source format for encoding chemical reactions.
• Provide a flexible and machine-readable representation for reactions.

Main Methods:

• Developed ReactionCode, an open-source format for encoding reactions.
• Represented reactions as a condensed graph of reaction (CGR), aggregating reactants and products.
• Designed the format for multi-layer machine readability.

Main Results:

• ReactionCode enables encoding and decoding of chemical reactions into a condensed graph representation.
• The format is flexible for various cheminformatics applications.
• Facilitates reaction similarity searching, classification, and database organization.

Conclusions:

• ReactionCode offers a standardized, open-source solution for reaction representation.
• The format supports advanced applications like machine learning and serves as a transform reaction language.