STRING is a database that integrates and transfers protein-protein associations across organisms. It combines experimental data, database mining, and computational predictions to provide a comprehensive collection of associations for a large number of organisms. The associations are ranked based on a common reference set and presented in a web interface. STRING extends associations beyond the original organism by transferring them to orthologous protein pairs in other organisms. The database contains 730,000 proteins in 180 fully sequenced organisms and is available at http://string.embl.de/. STRING integrates three types of evidence: direct experimental interactions, functional groupings, and predicted associations. It uses a scoring system to assess the confidence of each association, based on benchmarking against a reference set. The database also includes associations derived from the literature and functional genomics data. The transfer of associations between organisms is done using two strategies: COG-mode, which relies on orthology assignments, and protein-mode, which uses sequence similarity searches. STRING provides a combined score for each association, which is calculated based on the individual scores from different sources. This score reflects the confidence of the association when supported by multiple types of evidence. The database is an exploratory resource that offers a large number of associations with varying confidence scores, making it useful for quickly gaining an initial overview of the functional partners of a query protein.STRING is a database that integrates and transfers protein-protein associations across organisms. It combines experimental data, database mining, and computational predictions to provide a comprehensive collection of associations for a large number of organisms. The associations are ranked based on a common reference set and presented in a web interface. STRING extends associations beyond the original organism by transferring them to orthologous protein pairs in other organisms. The database contains 730,000 proteins in 180 fully sequenced organisms and is available at http://string.embl.de/. STRING integrates three types of evidence: direct experimental interactions, functional groupings, and predicted associations. It uses a scoring system to assess the confidence of each association, based on benchmarking against a reference set. The database also includes associations derived from the literature and functional genomics data. The transfer of associations between organisms is done using two strategies: COG-mode, which relies on orthology assignments, and protein-mode, which uses sequence similarity searches. STRING provides a combined score for each association, which is calculated based on the individual scores from different sources. This score reflects the confidence of the association when supported by multiple types of evidence. The database is an exploratory resource that offers a large number of associations with varying confidence scores, making it useful for quickly gaining an initial overview of the functional partners of a query protein.