APID (Agile Protein Interaction DataAnalyzer) 23 APID (Agile Protein Interaction DataAnalyzer) Integrates and unifies 7 DBs: BIND, DIP, HPRD, IntAct, MINT, BioGRID. Includes 51,873 proteins 241,204 interactions Includes quality assessment methods based on: - nº of experimental methods - functional similarity (GO) - interacting domains (ipfam) 24
APID (Agile Protein Interaction DataAnalyzer) Problems found in PPI data and DBs Quite heterogene (little overlap) Each DB take different scientific papers and have different curation teams Each DB follows different annotation protocols Still partial (incomplete ) Include many false positives (noisy) comparison done in 2006 25 APID (6 source DBs) BIND (University of Toronto) BioGRID (Mount Sinai Hospital) DIP (UCLA) HPRD (Johns Hopkins University) IntAct (EBI) 51,873 proteins 241,204 interactions MINT (University Rome)! 22,781 proteins! 50,218 interactions! 529,018 proteins???! 98,867 interactions! 19,935 proteins! 56,638 interactions! 25,661 proteins! 38,167 interactions! 63,568 proteins! 111,249 interactions! 28,817 proteins! 105,899 interactions NOTE: the numbers are in continuous revision by the DBs (look at their web sites). 26
Proteins per Organism: APID Statistics 51,873 proteins 241,204 interactions Interactions per Organism: ORGANISM (SPECIES) Homo sapiens Drosophila melanogaster NUMBER OF PROTEINS 12087 11483 ORGANISM (SPECIES) Saccharomyces cerevisiae Homo sapiens NUMBER OF INTERACTIONS 68843 60610 Saccharomyces cerevisiae Mus musculus Caenorhabditis elegans Escherichia coli Rattus norvegicus 5889 4008 3713 2935 1614 Drosophila melanogaster Escherichia coli Campylobacter jejuni Caenorhabditis elegans 47320 17864 11961 7684 Mus musculus 6228 27 APID Search Protein A protein name, protein identifier, protein description or part of it can be inserted in the general APID search tool. 28
APID Protein Query Results The figure shows the result given by the search for RASH. A table with 3 rows (one for each result) is presented. This table includes six columns with information about proteins: the UniProt entry name, the number of interactions, the UniProt_ID number, the taxon (NCBI Taxonomy ID), the protein name or description and a link to more information about the protein. 29 APID Protein More Information, +info_prot Clicking on +info_prot a new window with more detailed information about the query protein is displayed, including links to other referred biomolecular databases. The +info_prot file also includes some calculated parameters about the protein interaction network (i.e. connectivity and cluster coefficient) and about the protein functional environment based on GO annotation (i.e. GO environment enrichment) 30
APID Protein Interactions Clicking on the 81 link (interactions column) in the Protein Query Results a new page is displayed including a table with details about the 81 interactions that have been reported for RASH_HUMAN. This table has 5 columns with information about: the interaction protein partners, the number of experiments that validate each interaction, the provenance source databases (with links to them) and a final column with more information about the interaction: +info_inter 31 APID Interaction More Information, +info_inter Clicking on any +info_inter (in Protein Interaction page) a new window with more detailed information about the corresponding interaction protein pair is displayed, including marks in yellow that show GO terms overlapping and marks in green that show ipfam domain-domain interactions. 32
APID Interaction Filter Protein interactions can be selected using a filter that restrict the display to interaction pairs proven by 5 methods at least and that also show ipfam domain domain interactions. Doing this the number of interaction partners for RASH_HUMAN is reduced to only 6 proteins. 33 APID Experiments Clicking on the number of experiments APID displays another window with the information about all the experiments that validate any given PPI (e.g. RASH_HUMAN - SOS1_HUMAN), presenting for each experiment: (i) the publications that describe such interaction, linking to PubMed (by accession number PMID) and including a description of the publication; (ii) the type of method, linked to the publication that explains such experimental technique and the PSI-MI method-identifier; (iii) the source databases that include these data. 34
APID Network Browser APID also includes a Graph button that opens a graphical interactive network browser, where the proteins are nodes and the interactions edges. This application tool visualizes dynamically the data, and allows interactive exploring and navigating along the network. 35 How to submit your interaction data? 36
How to submit your interaction data? From Orchard et al. (2007) Nature Biotech. 37 How to submit your interaction data? From Orchard et al. (2007) 38