BioGRID

BioGRID (Biological General Repository for Interaction Datasets) is a curated open-access database of protein-protein interactions, genetic interactions, and chemical interactions, maintained by the University of Toronto. It is one of the foundational databases that consolidates experimental evidence from peer-reviewed literature into a standardized, searchable format, enabling researchers to query interaction networks for any gene or protein of interest. Unlike high-throughput interaction maps generated by yeast two-hybrid or mass spectrometry screens, BioGRID emphasizes low-throughput, individually validated interactions — a trade-off that sacrifices coverage for reliability.

The database assigns evidence codes to each interaction, distinguishing physical associations from genetic dependencies and providing the experimental system in which the interaction was detected. This curation strategy produces a network that is biased toward well-studied proteins and pathways, because highly funded and historically prominent research areas generate more publications and therefore more curated interactions. BioGRID thus encodes not only molecular biology but also the sociology of scientific attention.

BioGRID integrates with larger network analysis platforms such as Cytoscape and the Gene Ontology, and its data underpins many network medicine and systems biology studies. The challenge for the next generation of biological databases is not merely to add more interactions but to weight them by reliability, context, and functional relevance — a problem that requires moving beyond Boolean interaction graphs to probabilistic or condition-specific models.

BioGRID and similar databases are treated as ground truth by researchers who build network models, but they are better understood as curated hypotheses. Every "interaction" in BioGRID is an interpretation of an experiment, filtered through curation rules, publication bias, and the conceptual framework of the curators. The day we stop treating databases as raw data and start treating them as theory-laden instruments is the day network biology becomes a mature science.