3.5: Select prediction method to search for interacting proteins:
PeptideMine server is using Protein-Protein Interactions recorded in STRING database in the back-end to identify known and predicted interacting partners of a given protein. Users can control the searches to identify interacting partners by selecting different prediction methods used in the STRING. Users are encouraged to refer to the STRING help page and blog about the STRING for more details. We are providing 8 different methods available from the STRING database. Users can select any one of the 7 different prediction methods(Neighborhood, Gene Fusion, Co-expression, Experiments, Co-occurence, Databases, Textmining) or 'All method'(combined method of all 7 prediction methods) to identify the interaction. Interacting partners of a given gene is identified by searching the direct interacting partners (level 1, only binary interactions are considered in the current version) based on the selected Confidence score and prediction method. The search is performed in the local version of STRING database in the backend of PeptideMine server.