The article introduces STRING, a comprehensive database that integrates and ranks protein-protein associations across various organisms. STRING aims to provide a unified framework for direct experimental evidence, functional grouping, and predicted associations. The database includes 730,000 proteins in 180 fully sequenced organisms and is accessible through a web interface. Protein associations are derived from high-throughput experiments, database mining, literature extraction, and genomic context analysis. Confidence scores are assigned to each association, and interactions are transferred between organisms using two methods: COG-mode, which relies on orthology assignments, and protein-mode, which uses sequence similarity. The final combined score reflects the confidence of the association, combining evidence from multiple sources. STRING is designed to facilitate comparative analysis, increase coverage, and support evolutionary studies.The article introduces STRING, a comprehensive database that integrates and ranks protein-protein associations across various organisms. STRING aims to provide a unified framework for direct experimental evidence, functional grouping, and predicted associations. The database includes 730,000 proteins in 180 fully sequenced organisms and is accessible through a web interface. Protein associations are derived from high-throughput experiments, database mining, literature extraction, and genomic context analysis. Confidence scores are assigned to each association, and interactions are transferred between organisms using two methods: COG-mode, which relies on orthology assignments, and protein-mode, which uses sequence similarity. The final combined score reflects the confidence of the association, combining evidence from multiple sources. STRING is designed to facilitate comparative analysis, increase coverage, and support evolutionary studies.