STRING is a database of known and predicted protein-protein interactions.
The interactions include direct (physical) and indirect (functional) associations. STRING quantitatively integrates interaction data from various sources for a large number of organisms, and transfers information between these organisms where applicable. The database currently contains 1,513,782 proteins in 373 species. This program allows you to download a subgraph composed of the neighbours of some query genes in the STRING database.
For more information on STRING, see Von Mering et al, 2007