TM-IMC

TM-IMC predicts Gene Ontology (GO) term associations for G protein-coupled receptors (GPCRs) by integrating text mining and inductive matrix completion.

Key Features:

• Text Mining: Extracts functional information from GPCR-associated biomedical literature using text-mining techniques.
• Inductive Matrix Completion: Applies inductive matrix completion models to predict associations between GPCRs and GO terms, addressing molecular function and biological process annotations.
• Large-Scale Benchmarking: Evaluates prediction performance through large-scale benchmarking against baseline protein function annotation and literature-based GO annotation methods.

Scientific Applications:

• GO Term Annotation for GPCRs: Predicts Gene Ontology annotations for G protein-coupled receptors to support functional characterization.
• Molecular Function and Biological Process Annotation: Improves assignment of molecular function and biological process GO terms to GPCR proteins.
• Literature-Driven Functional Discovery: Enables systematic, literature-based identification of putative GPCR functions for downstream experimental study.

Methodology:

Performs a three-stage approach that extracts GPCR-associated information from biomedical literature via text mining, applies inductive matrix completion to predict GPCR–GO term associations, and assesses results with large-scale benchmarking against baseline methods.