Theoretical and Computational Chemistry

Chemical Reaction Network Knowledge Graphs: the OntoRXN Ontology

Authors

  • Diego Garay-Ruiz Institute of Chemical Research of Catalonia (ICIQ) & Rovira i Virgili University ,
  • Carles Bo Institute of Chemical Research of Catalonia (ICIQ) & Rovira i Virgili University

Abstract

The organization and management of large amounts of data has become a major point in almost all areas of human knowledge. In this context, semantic approaches propose a structure for the target data, defining ontologies that state the types of entities on a certain field and how these entities are interrelated. In this work, we introduce OntoRXN, a novel ontology describing the reaction networks constructed from computational chemistry calculations. Under our paradigm, these networks are handled as undirected graphs, without assuming any traversal direction. From there, we propose a core class structure including reaction steps, network stages, chemical species, and the lower-level entities for the individual computational calculations. These individual calculations are founded on the OntoCompChem ontology and on the ioChem-BD database, where information is parsed and stored in CML format. OntoRXN is introduced through several examples in which knowledge graphs based on the ontology are generated for different chemical systems available on ioChem-BD. Finally, the resulting knowledge graphs are explored through SPARQL queries, illustrating the power of the semantic approach to standardize the analysis of intricate datasets and to simplify the development of complex workflows.

Content

Thumbnail image of OntoRXN_Draft.pdf

Supplementary material

Thumbnail image of OntoRXN_SI.pdf
Supporting Info: Chemical Reaction Network Knowledge Graphs: the OntoRXN Ontology
Supporting Information, including additional details on some of the workflows mentioned along the main text

Supplementary weblinks

OntoRXN repository
GitLab repository including ontology specification
OntoRXN-Tools repository
GitLab repository including knowledge graph instantiation library.