The Microbial Protein Interaction Database
(1) Enter a common gene name, e.g. cheY or an ordered locus tag 'TM_0700' or a UniProt accession 'Q56312'
(2) Enter a species name e.g. 'Themotoga maritima' or a part of a species name 'maritima'
(3) Species for which MPIDB stores interaction data can be selected from the drop down menu.
(4) Select interactions based on source datasets, e.g. MPIDB's literature curated dataset MPI-LIT.
(5) Select interactions based on the experimental method defined by the PSI MI ontology.
(6) Select interactions based on the minimum number of 3D structures which contain both interacting partners.
(7) Select interactions based on the minimum number of different PSI MI methods.
(8) Select interactions based on the minimum number of conserved interactions (interologs).
(9) Select interactions based on the minimum number of co-purifications of interacting proteins.
(10) Select interactions based on the minimum number of overall evidences (Sum of experimental, interolog, co-purification, domain-domain contact, and secondary references evidences.
(11) Click to start the search.
(1) Number of found interactions.
(2) Click to download found interactions.
(3) Interactions are sorted descendingly on the number of interaction evidences. Clicking on the number of evidences shows experimental and predicted evidences along with links to the source databases (see next section).
(4) Page navigation.
(1) Interacting protein names with species/strain information.
(2) Thumbnail of a three-dimensional structure. Hover over it to enlarge it. Click on the structure to link to the Molecular Structure Database (MSD) at the EBI.Thumbnails are displayed whenever both interacting proteins are part of a structural assembly (multimeric structure of at least two proteins). For homodimeric interactions this feature is not available.
(4) Summary of evidence that support the interaction (here tsf - tufB). Evidences are either experimental or predictive. Interactions are supported by at least one experimental interaction evidence.
(5) Number of Experimental Evidences. Experimental evidences are based on experiments that are designed to detect a physical contact between two proteins. An evidence is defined by its experimental interaction detection method (PSI-MI controlled vocabulary) in combination with its publication (PubMed ID). Experiments from pull down studies were included onlu if a bait protein identified only one prey protein.
(6) Number of Interologs. Interologs have been determined by sequence comparison of interacting proteins (BLAST E-value cut off <=1.0E-5).
(7) Interolog example: the left protein (tsf of Thermus thermophilus) is a sequence homolog of the upper left protein (tsf of Escherichia coli). The same is true for the right protein (tufA of Thermus thermophilus and tufB of Escherichia coli). The percentage of sequence conservation for each protein is given behind its name. E.g. 21% of amino acids of the tsf (E. coli) protein are identical with amino acids from tsf (Thermus thermophilus). The geometric mean of BLAST identities, positives and E-values of the two pairwise sequence comparisons are given below. Thumbnails of structures are displayed whenever both homologous proteins are part of a 3D structural assembly. Interologs are sorted descending by the geometric mean of pairwise percent identities
(8) Interaction link: click on this symbol to show an overview of interaction evidences for the respective interolog (here tsf - tufB (Thermus thermophilus)).
(9) Experimental evidences for a specific interolog (here tsf - tufB (Thermus thermophilus)): the first column indicates the experimental method according to the PSI MI ontology.
(10) Source Database links: a click on the database logo links to the source database that describes this experiment.
(11) For each experiment the source publication can be looked up in PubMed by clicking on the first author's name.
(12) For interactions which also have been predicted from complex purification data (matrix model), complex purification evidences are displayed. The type field indicates that the evidence is predictive (the exact topology of complex members is unknown). It also contains information on the size of the complex. Evidences from co-purifications are ordered ascending by their number of purified proteins.
(13) Domain-Domain Interactions: Pfam domain combinations that have been found to form close contacts in three-dimensional structures (iPfam, 3DID). Here, three Pfam domain-domain combination have been found among the interacting proteins that are also reported by iPfam and/or 3did. Domains are linked with the Pfam database.
(14) Secondary References: list of publications that have been cited for a certain interaction (such publications have not been followed up to extract experimental descriptions and therefore have been included as predicted evidences).
(15) Link to a contact form for questions and comments on an interaction (here tsf - tufB).