GlyS3

GlyS3 performs glycan substructure searches across structural databases (UniCarbKB, SugarBind, GlycomeDB) using graph representations and graph query languages to enable analysis of glycan structural determinants.

Key Features:

• Graph Database Utilization: Models glycan structures as direct acyclic graphs stored in RDF triple stores and Property Graph databases where nodes represent glycan building blocks and edges denote chemical linkages.
• Substructure Search Capability: Executes substructure searches to retrieve specific glycan regions and identify determinants recognized by glycan-binding proteins (GBPs).
• Query Languages Support: Supports SPARQL for RDF and Cypher for Property Graph implementations to perform precise graph-structured queries.

Scientific Applications:

• GBP recognition mapping: Identification of glycan determinants recognized by glycan-binding proteins (GBPs) from structural databases.
• Cellular communication and immune studies: Analysis of glycan structural motifs relevant to cellular communication and immune response modulation.
• Disease-associated glycosylation analysis: Investigation of glycosylation pattern changes linked to disease pathogenesis.

Methodology:

Modeled 19,404 glycan structures from GlycomeDB for storage in RDF triple stores and Property Graph databases, and evaluated two sets of substructure searches comparing query response times and result equivalence between the two database technologies.