Search

CancerGeneNet (https://signor.uniroma2.it/CancerGeneNet/) is a resource that links genes that are frequently mutated in cancers to cancer phenotypes. The resource takes advantage of a curation effort aimed at embedding a large fraction of the gene products that are found altered in cancer cells into a network of causal protein relationships. Graph algorithms, in turn, allow to infer likely paths of causal interactions linking cancer associated genes to cancer phenotypes thus offering a rational framework for the design of strategies to revert disease phenotypes. CancerGeneNet bridges two interaction layers by connecting proteins whose activities are affected by cancer drivers to proteins that impact on the 'hallmarks of cancer'. In addition, CancerGeneNet annotates curated pathways that are relevant to rationalize the pathological consequences of cancer driver mutations in selected common cancers and 'MiniPathways' illustrating regulatory circuits that are frequently altered in different cancers., (© The Author(s) 2019. Published by Oxford University Press on behalf of Nucleic Acids Research.)

The International Molecular Exchange (IMEx) consortium is an international collaboration between major public interaction data providers to share literature-curation efforts and make a nonredundant set of protein interactions available in a single search interface on a common website (http://www.imexconsortium.org/). Common curation rules have been developed, and a central registry is used to manage the selection of articles to enter into the dataset. We discuss the advantages of such a service to the user, our quality-control measures and our data-distribution practices.

Background: Determining usefulness of biomedical text mining systems requires realistic task definition and data selection criteria without artificial constraints, measuring performance aspects that go beyond traditional metrics. The BioCreative III Protein-Protein Interaction (PPI) tasks were motivated by such considerations, trying to address aspects including how the end user would oversee the generated output, for instance by providing ranked results, textual evidence for human interpretation or measuring time savings by using automated systems. Detecting articles describing complex biological events like PPIs was addressed in the Article Classification Task (ACT), where participants were asked to implement tools for detecting PPI-describing abstracts. Therefore the BCIII-ACT corpus was provided, which includes a training, development and test set of over 12,000 PPI relevant and non-relevant PubMed abstracts labeled manually by domain experts and recording also the human classification times. The Interaction Method Task (IMT) went beyond abstracts and required mining for associations between more than 3,500 full text articles and interaction detection method ontology concepts that had been applied to detect the PPIs reported in them., Results: A total of 11 teams participated in at least one of the two PPI tasks (10 in ACT and 8 in the IMT) and a total of 62 persons were involved either as participants or in preparing data sets/evaluating these tasks. Per task, each team was allowed to submit five runs offline and another five online via the BioCreative Meta-Server. From the 52 runs submitted for the ACT, the highest Matthew's Correlation Coefficient (MCC) score measured was 0.55 at an accuracy of 89% and the best AUC iP/R was 68%. Most ACT teams explored machine learning methods, some of them also used lexical resources like MeSH terms, PSI-MI concepts or particular lists of verbs and nouns, some integrated NER approaches. For the IMT, a total of 42 runs were evaluated by comparing systems against manually generated annotations done by curators from the BioGRID and MINT databases. The highest AUC iP/R achieved by any run was 53%, the best MCC score 0.55. In case of competitive systems with an acceptable recall (above 35%) the macro-averaged precision ranged between 50% and 80%, with a maximum F-Score of 55%., Conclusions: The results of the ACT task of BioCreative III indicate that classification of large unbalanced article collections reflecting the real class imbalance is still challenging. Nevertheless, text-mining tools that report ranked lists of relevant articles for manual selection can potentially reduce the time needed to identify half of the relevant articles to less than 1/4 of the time when compared to unranked results. Detecting associations between full text articles and interaction detection method PSI-MI terms (IMT) is more difficult than might be anticipated. This is due to the variability of method term mentions, errors resulting from pre-processing of articles provided as PDF files, and the heterogeneity and different granularity of method term concepts encountered in the ontology. However, combining the sophisticated techniques developed by the participants with supporting evidence strings derived from the articles for human interpretation could result in practical modules for biological annotation workflows.

Many fundamental processes involve protein-protein interactions. Recent advances in technology make it possible to perform large-scale, genome-wide interaction mapping experiments that result in an always increasing amount of data. Protein-protein interaction databases are thus becoming a major resource for investigating biological networks and pathways. In this chapter we describe the Molecular INTeraction database (MINT). The MINT database aims at storing, in a structured format, information about protein-protein interactions (PPIs) by extracting experimental details from work published in peer-reviewed journals.

The study of the complex web of interactions that link biological molecules in a cell is the subject of interactomics--currently one of the fastest moving fields in molecular biology. The recent completion of high-throughput studies to investigate systematically all the possible interactions in a variety of model organisms has provided unique opportunities to compare interaction networks and ask questions about their conservation during evolution. It is expected that this approach will yield a scientific return as rich as that obtained in the past decade from comparing genomes and proteomes from different organisms.

Background: The application of high throughput approaches to the identification of protein interactions has offered for the first time a glimpse of the global interactome of some model organisms. Until now, however, such genome-wide approaches have not been applied to the human proteome., Results: In order to fill this gap we have assembled an inferred human protein interaction network where interactions discovered in model organisms are mapped onto the corresponding human orthologs. In addition to a stringent assignment to orthology classes based on the InParanoid algorithm, we have implemented a string matching algorithm to filter out orthology assignments of proteins whose global domain organization is not conserved. Finally, we have assessed the accuracy of our own, and related, inferred networks by benchmarking them against i) an assembled experimental interactome, ii) a network derived by mining of the scientific literature and iii) by measuring the enrichment of interacting protein pairs sharing common Gene Ontology annotation., Conclusion: The resulting networks are named HomoMINT and HomoMINT_filtered, the latter being based on the orthology table filtered by the domain architecture matching algorithm. They contains 9749 and 5203 interactions respectively and can be analyzed and viewed in the context of the experimentally verified interactions between human proteins stored in the MINT database. HomoMINT is constantly updated to take into account the growing information in the MINT database.

A substantial fraction of protein interactions in the cell is mediated by families of protein modules binding to relatively short linear peptides. Many of these interactions have a high dissociation constant and are therefore suitable for supporting the formation of dynamic complexes that are assembled and disassembled during signal transduction. Extensive work in the past decade has shown that, although member domains within a family have some degree of intrinsic peptide recognition specificity, the derived interaction networks display substantial promiscuity. We review here recent advances in the methods for deriving the portion of the protein network mediated by these domain families and discuss how specific biological outputs could emerge in vivo despite the observed promiscuity in peptide recognition in vitro.

A major goal of proteomics is the complete description of the protein interaction network underlying cell physiology. A large number of small scale and, more recently, large-scale experiments have contributed to expanding our understanding of the nature of the interaction network. However, the necessary data integration across experiments is currently hampered by the fragmentation of publicly available protein interaction data, which exists in different formats in databases, on authors' websites or sometimes only in print publications. Here, we propose a community standard data model for the representation and exchange of protein interaction data. This data model has been jointly developed by members of the Proteomics Standards Initiative (PSI), a work group of the Human Proteome Organization (HUPO), and is supported by major protein interaction data providers, in particular the Biomolecular Interaction Network Database (BIND), Cellzome (Heidelberg, Germany), the Database of Interacting Proteins (DIP), Dana Farber Cancer Institute (Boston, MA, USA), the Human Protein Reference Database (HPRD), Hybrigenics (Paris, France), the European Bioinformatics Institute's (EMBL-EBI, Hinxton, UK) IntAct, the Molecular Interactions (MINT, Rome, Italy) database, the Protein-Protein Interaction Database (PPID, Edinburgh, UK) and the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING, EMBL, Heidelberg, Germany).

IntAct provides an open source database and toolkit for the storage, presentation and analysis of protein interactions. The web interface provides both textual and graphical representations of protein interactions, and allows exploring interaction networks in the context of the GO annotations of the interacting proteins. A web service allows direct computational access to retrieve interaction networks in XML format. IntAct currently contains approximately 2200 binary and complex interactions imported from the literature and curated in collaboration with the Swiss-Prot team, making intensive use of controlled vocabularies to ensure data consistency. All IntAct software, data and controlled vocabularies are available at http://www.ebi.ac.uk/intact.

The Molecular Interactions Database (MINT) is a relational database designed to store information about protein interactions. Expert curators extract the relevant information from the scientific literature and deposit it in a computer readable form. Currently (April 2003) MINT contains information on 4300 proteins and >3000 interactions. Although MINT contains interactions among proteins for several different species, the curators have initially focused on protein interactions in mammalian species. Protocols are provided here for searching MINT over the Internet, using the MINT Viewer.

We propose a normalization of symbols and terms used to describe, accurately and succinctly, the detailed interactions between amino acid residues of pairs of interacting proteins at protein:protein (or protein:peptide) interfaces. Our aim is to unify several diverse descriptions currently in use in order to facilitate communication in the rapidly progressing field of signaling by protein domains. In order for the nomenclature to be convenient and widely used, we also suggest a parallel set of symbols restricted to the ASCII format allowing accurate parsing of the nomenclature to a computer-readable form. This proposal will be reviewed in the future and will therefore be open for the inclusion of new rules, modifications and changes.

Protein interaction domain families that modulate the formation of macromolecular complexes recognize specific sequence or structural motifs. For instance SH3 and WW domains bind to polyproline peptides while SH2 and FHA domains bind to peptides phosphorylated in Tyr and Thr respectively. Within each family, variations in the chemical characteristics of the domain binding pocket modulate a finer peptide recognition specificity and, as a consequence, determine the selection of functional protein partners in vivo. In the proteomic era there is the need for reliable inference methods to help restricting the sequence space of the putative targets to be confirmed experimentally by more laborious experimental approaches. Here we will review the published data about the peptide recognition specificity of the SH3 domain family and we will propose a classification of SH3 domains into eight classes. Finally, we will discuss whether the available information is sufficient to infer the recognition specificity of any uncharacterized SH3 domain.

Protein interaction databases represent unique tools to store, in a computer readable form, the protein interaction information disseminated in the scientific literature. Well organized and easily accessible databases permit the easy retrieval and analysis of large interaction data sets. Here we present MINT, a database (http://cbm.bio.uniroma2.it/mint/index.html) designed to store data on functional interactions between proteins. Beyond cataloguing binary complexes, MINT was conceived to store other types of functional interactions, including enzymatic modifications of one of the partners. Release 1.0 of MINT focuses on experimentally verified protein-protein interactions. Both direct and indirect relationships are considered. Furthermore, MINT aims at being exhaustive in the description of the interaction and, whenever available, information about kinetic and binding constants and about the domains participating in the interaction is included in the entry. MINT consists of entries extracted from the scientific literature by expert curators assisted by 'MINT Assistant', a software that targets abstracts containing interaction information and presents them to the curator in a user-friendly format. The interaction data can be easily extracted and viewed graphically through 'MINT Viewer'. Presently MINT contains 4568 interactions, 782 of which are indirect or genetic interactions.

Peptide recognition modules mediate many protein-protein interactions critical for the assembly of macromolecular complexes. Complete genome sequences have revealed thousands of these domains, requiring improved methods for identifying their physiologically relevant binding partners. We have developed a strategy combining computational prediction of interactions from phage-display ligand consensus sequences with large-scale two-hybrid physical interaction tests. Application to yeast SH3 domains generated a phage-display network containing 394 interactions among 206 proteins and a two-hybrid network containing 233 interactions among 145 proteins. Graph theoretic analysis identified 59 highly likely interactions common to both networks. Las17 (Bee1), a member of the Wiskott-Aldrich Syndrome protein (WASP) family of actin-assembly proteins, showed multiple SH3 interactions, many of which were confirmed in vivo by coimmunoprecipitation.

Several approaches, some of which are described in this issue, have been proposed to assemble a complete protein interaction map. These are often based on high throughput methods that explore the ability of each gene product to bind any other element of the proteome of the organism. Here we propose that a large number of interactions can be inferred by revealing the rules underlying recognition specificity of a small number (a few hundreds) of families of protein recognition modules. This can be achieved through the construction and characterization of domain repertoires. A domain repertoire is assembled in a combinatorial fashion by allowing each amino acid position in the binding site of a given protein recognition domain to vary to include all the residues allowed at that position in the domain family. The repertoire is then searched by phage display techniques with any target of interest and from the primary structure of the binding site of the selected domains one derives rules that are used to infer the formation of complexes between natural proteins in the cell.

We have developed a procedure to predict the peptide binding specificity of an SH3 domain from its sequence. The procedure utilizes information extracted from position-specific contacts derived from six SH3/peptide or SH3/protein complexes of known structure. The framework of SH3/peptide contacts defined on the structure of the complexes is used to build a residue-residue interaction database derived from ligands obtained by panning peptide libraries displayed on filamentous phage. The SH3-specific interaction database is a multidimensional array containing frequencies of position-specific contacts. As input, SH3-SPOT requires the sequence of an SH3 domain and of a query decapeptide ligand. The array, that we call the SH3-specific matrix, is then used to evaluate the probability that the peptide would bind the given SH3 domain. This procedure is fast enough to be applied to the entire protein sequence database. Panning experiments were performed to search putative specific ligands of different SH3 domains in a database of decapeptides, or in a database of protein sequences. The procedure ranked some of the natural partners of interaction of a number of SH3 domains among the best ligands of the approximately 5. 6x10(9) different decapeptides in the SWISSPROT database. We expect the predictive power of the method to increase with the enrichment of the SH3-specific matrix by interaction data derived from new complex structures or from the characterization of new ligands. The procedure was developed using the SH3 domain family as test case but its application can easily be extended to other families of protein domains (such as, SH2, MHC, EH, PDZ, etc.)., (Copyright 2000 Academic Press.)

SH3 domains constitute a family of protein-protein interaction modules that bind to peptides displaying an X-proline-X-X-proline (XPXXP) consensus. We report that the SH3 domain of Eps8, a substrate of receptor and non-receptor tyrosine kinases, displays a novel and unique binding preference. By a combination of approaches including (i) screening of phage-displayed random peptide libraries, (ii) mapping of the binding regions on three physiological interactors of Eps8, (iii) alanine scanning of binding peptides and (iv) in vitro cross-linking, we demonstrate that a proline-X-X-aspartate-tyrosine (PXXDY) consensus is indispensable for binding to the SH3 domain of Eps8. Screening of the Expressed Sequence Tags database allowed the identification of three Eps8-related genes, whose SH3s also display unusual binding preferences and constitute a phylogenetically distinct subfamily within the SH3 family. Thus, Eps8 identifies a novel family of SH3-containing proteins that do not bind to canonical XPXXP-containing peptides, and that establish distinct interactions in the signaling network.

We report a procedure for the description and comparison of protein surfaces, which is based on a three-dimensional (3D) transposition of the profile method for sensitive protein homology sequence searches. Although the principle of the method can be applied to detect similarities to a single protein surface, the possibility of extending this approach to protein families displaying common structural and/or functional properties, makes it a more powerful tool. In analogy to profiles derived from the multiple alignment of protein sequences, we derive a 3D surface profile from a protein structure or from a multiple structure alignment of several proteins. The 3D profile is used to screen the protein structure database, searching for similar protein surfaces. The application of the procedure to SH2 and SH3 binding pockets and to the nucleotide binding pocket associated with the p-loop structural motif is described. The SH2 and SH3 3D profiles can identify all the SH2 and SH3 binding regions present in the test dataset; the p-loop 3D profile is able to recognize all the p-loop-containing proteins present in the test dataset. Analysis of the p-loop 3D profile allowed the identification of a positive charge whose position is conserved in space but not in sequence. The best ranking non-p-loop-containing protein is an ADP-forming succinyl coenzyme A synthetase, whose nucleotide-binding region has not yet been identified., (Copyright 1998 Academic Press.)

The synapsins are a family of major neuron-specific synaptic vesicle-associated phosphoproteins which play important roles in synaptic function. In an effort to identify molecular tools which can be used to perturb the activity of the synapsins in in vitro as well as in vivo experiments, we have localized the epitopes of a panel of monoclonal antibodies (mAbs) raised against synapsins I and II and have characterized their ability to interfere with the interactions of the synapsins with protein kinases, actin and Src homology-3 (SH3) domains. The epitopes of the six mAbs were found to be concentrated in the N-terminal region within domains A and B for the synapsin II-reactive mAbs 19.4, 19.11, 19.51 and 19.21, and in two C-terminal clusters in the proline-rich domains D for synapsin I (mAbs 10.22, 19.51, 19.11 and 19.8) and G for synapsin II (mAb 19.8). The synapsin II-specific mAbs 19.4 and 19.21, whose overlapping epitopes are adjacent to phosphorylation site 1, specifically inhibited synapsin II phosphorylation by endogenous or exogenous cAMP-dependent protein kinase. While all the anti-synapsin I mAbs were unable to affect the interactions of synapsin I both with Ca2+/calmodulin-dependent protein kinase II and with actin monomers and filaments, mAbs 19.8 and 19.51 were found to inhibit the binding of Grb2 SH3 domains to the proline-rich C-terminal region of synapsin I.

Shc proteins are targets of activated tyrosine kinases and have been implicated in the transmission of activation signals to Ras. Upon phosphorylation, Shc proteins form stable complexes with cellular tyrosine-phosphorylated proteins and with the Grb2 adaptor protein. Two Shc isoforms of 52 and 46 kDa have been characterized. They share a C-terminal SH2 domain, a proline- and glycine-rich region (collagen homologous region 1; CH1) and a N-terminal phospho-tyrosine binding domain (PTB). We report her ethe initial characterization of two Shc related human cDNAs: ShcB and ShcC. The ShcB and ShcC cDNAs code for proteins that are highly similar and share the same modular organization as Shc. PTB and SH2 domains of ShcB and ShcC have similar binding specificities in vitro and bind to activated EGFR in a phosphotyrosine-dependent manner. Based on these findings we propose to rename Shc as ShcA. Anti-ShcB and anti-ShcC antibodies recognize specific polypeptides of 52, 47 kDa (ShcB) and 54 kDa (ShcC) in mammalian cells. Since these two genes are predominantly expressed in specific brain tissues, these Shc family members may be involved in cell type-specific signaling, in the nervous system.