Description of PPI databases and repositories.

The table divided in three sections: primary databases, which include PPIs from large- and small-scale (Lsc & Ssc) experimental data that are usually obtained from curation of research articles (8 resources included: BIND, BioGRID, DIP, HPRD, IntAct, MINT, MIPS-MPACT, MIPS-MPPI); meta-databases, which include PPIs derived from integration and unification of several primary repositories (3 resources: APID, MPIDB, PINA); prediction databases, which include PPIs from experimental analyses together with predicted PPIs obtained from the analyses of heterogenous biological data (5 resources: MiMI, PIPs, OPHID, STRING, UniHI). The table shows the total number of proteins and interactions that were reported by each repository in December 2009 (as far as we could see in the respective Web site). The numbers are in brackets [ ] when the repository includes PPIs and other types of interactions (e.g., protein-ligand interactions or for the case of prediction databases nonPPI data). The question mark [?] indicates that the number of distinct proteins included is such repository could not be found in the Web.