STRING v9.1 is an updated version of the protein-protein interaction database that provides a more comprehensive and integrated view of protein interactions across a wide range of organisms. The database includes known and predicted associations, resulting in extensive protein networks covering over 1100 organisms. The update introduces several improvements, including enhanced text mining capabilities that now include full-text articles, a redesigned algorithm for transferring interactions between organisms, and the provision of statistical information on functional enrichment in networks. The text mining pipeline has been redesigned to improve the recognition of protein names and to use a new scoring scheme that considers co-occurrences within sentences, paragraphs, and documents. This has significantly improved the quality and number of associations extracted from text mining. The new system also incorporates orthology information from the eggNOG database to better integrate evidence across different organisms. The interaction transfer between organisms has been improved by using a hierarchical orthologous group approach based on the eggNOG database. This allows for more accurate transfer of functional associations between organisms by considering the phylogeny of organisms and their orthologous groups. The transfer process involves combining individual protein-protein links into links between orthologous groups and then transferring these links back to the protein level. The database also includes a new statistical enrichment analysis tool that allows users to explore whether their input protein lists show evidence of enrichment in known biological functions or pathways. This tool uses various functional annotation spaces, including Gene Ontology, KEGG, Pfam, and InterPro, and provides interactive visualization of the results. The user interface has been enhanced to allow users to log in, browse their search history, save pages, and upload protein lists of interest. Users can also provide additional data (payload information) that can be projected onto the network for further analysis. The database is designed to be user-friendly and comprehensive, providing a global perspective on protein interactions and their functional associations.STRING v9.1 is an updated version of the protein-protein interaction database that provides a more comprehensive and integrated view of protein interactions across a wide range of organisms. The database includes known and predicted associations, resulting in extensive protein networks covering over 1100 organisms. The update introduces several improvements, including enhanced text mining capabilities that now include full-text articles, a redesigned algorithm for transferring interactions between organisms, and the provision of statistical information on functional enrichment in networks. The text mining pipeline has been redesigned to improve the recognition of protein names and to use a new scoring scheme that considers co-occurrences within sentences, paragraphs, and documents. This has significantly improved the quality and number of associations extracted from text mining. The new system also incorporates orthology information from the eggNOG database to better integrate evidence across different organisms. The interaction transfer between organisms has been improved by using a hierarchical orthologous group approach based on the eggNOG database. This allows for more accurate transfer of functional associations between organisms by considering the phylogeny of organisms and their orthologous groups. The transfer process involves combining individual protein-protein links into links between orthologous groups and then transferring these links back to the protein level. The database also includes a new statistical enrichment analysis tool that allows users to explore whether their input protein lists show evidence of enrichment in known biological functions or pathways. This tool uses various functional annotation spaces, including Gene Ontology, KEGG, Pfam, and InterPro, and provides interactive visualization of the results. The user interface has been enhanced to allow users to log in, browse their search history, save pages, and upload protein lists of interest. Users can also provide additional data (payload information) that can be projected onto the network for further analysis. The database is designed to be user-friendly and comprehensive, providing a global perspective on protein interactions and their functional associations.