IDSM

IDSM integrates RDF-exported small-molecule data using ontologies and provides a SPARQL endpoint and structure-aware search capabilities to enable interoperable querying across chemical data resources.

Key Features:

• SPARQL Query Support: Supports SPARQL query language including federated queries to combine endpoints from independent data sources.
• Integration of Major Datasets: Integrates PubChem, ChEMBL, and ChEBI RDF datasets into a unified database accessible via a SPARQL endpoint.
• In-house SPARQL Engine: Implements a custom SPARQL engine that translates SPARQL queries into SQL and executes them on PostgreSQL with query optimizations.
• Enhanced Search Capabilities: Provides compound substructure and similarity searches implemented via the Sachem project for structure-driven chemical searches.

Scientific Applications:

• Drug design: Enables interoperable searches across chemical, protein, and assay data to support target and lead discovery workflows.
• Metabolomics: Facilitates cross-referencing of metabolites and related chemical information across integrated datasets.
• Cross-database linkage: Supports linking of chemicals to proteins, metabolites, and assays using semantic RDF representations and SPARQL queries.
• Structure-driven data discovery: Enhances identification of chemically similar compounds and substructures across large chemical repositories.

Methodology:

RDF-exported datasets are integrated into a unified database; a custom SPARQL engine translates SPARQL to SQL for execution on PostgreSQL and supports federated queries; Sachem provides compound substructure and similarity search functionality.