ProteINSIDE: a web service to computerize species similarity, gene ontology annotation, prediction of secreted peptide and protein-protein interactions


ProteINSIDE

ProteINSIDE, a novel and free web service, provides a systematic and integrative analysis of a protein’s biological data using (i) an overview of the biological information stored in public databases or provided by annotations according to the Gene Ontology (GO), (ii) the prediction of protein secretion to search for proteins that mediate signalisation between cells or tissues, and (iii) the identification and the analysis of protein-protein interactions (PPi) to identify proteins contributing to a process or to predict protein function. Using lists of proteins or genes as a unique input, ProteINSIDE is an original all-in-one tool that merges these three approaches to present a fast overview and comprehensive analysis of genomic data from Human, Rat, Murine, Bovine, Ovine, and Caprine species.


This pipeline can be use by this simple interface and there two choise of studies: a basical analysis and analysis fully customizable. All results are clearly readable and downloadable (like a "text" formatted file, pictures or Excels files).


Analyse Description Average time for 100 proteins (min)
Basic The basic analysis runs ProteINSIDE with automatic settings. It extracts the IDs and GO terms (without electronic annotation), predicts the proteins that could be secreted, and identifies the protein-protein interactions between the proteins on the dataset.

Settings for a basic analysis are ID Mapping / Gene Ontology / Signal Peptide prediction / Protein-protein Interaction research using three databases (IntAct / UniProtKB / BioGrid).
2 minutes
Custom The custom analysis runs ProteINSIDE with the settings requested by the user. Users can choose the selection of GO to include IEA codes and make GOTree chart networks with Cytoscape web. Users can select a specific database for PPi analysis, analyse protein-protein interaction in other species using protein data from orthologs and expand the network by looking for all the interactions known for proteins in the dataset. Users can choose a different sensitivity to detect signal peptides with SignalP 4.1.

Settings for a custom analysis depend of user's choices.
Settings for a fully custom analysis are ID Mapping / Gene Ontology -- GOTree -- Electronic annotation active / SignalP -- Sensitivity / Interaction research known on all PPi databases -- PPi with proteins in the dataset and external proteins /
Depends of user settings. A full analysis (all settings on) takes approximately 15 minutes

For better performance we recommand to use Google Chrome or Firefox or Internet Explorer 9+ as internet browser.


Softwares informations :

Software Informations Outputs Sources
Biological retrieval This software is completely invisible to the user; it performs queries on our database that merge data from UniprotKB and NCBI-Gene to retrieve the maximum biological information about proteins in the dataset. Biological information includes identifiers (genes, accession number, gi, protein ID), function, chromosome location, orthologs, functional annotation, and protein sequences (FASTA file).

Excel file / FASTA file

UniprotKB / NCBI-Gene

Ontologies This program selects GO from the QuickGO server for each protein (GO identifier, name, function and functional group). The selection depends on the evidence code (IEA codes are unselected in basic analysis, but can be selected in custom analysis) and the curator review of the annotation. It provides over- and under-represented terms to identify the most relevant, specific terms associated with the uploaded list. Within a GO, the script compares the number of genes or proteins from the dataset to the total number of gene products declared in the AmiGO database to provide a coverage frequency, and thus to identify the most common pathways associated to a dataset.

Diagrams / Excel files / GO Tree network

Quick GO / AmiGO / UniprotKB

Peptide Prediction This program runs a local version of SignalP 4.1. It uses the sequences of input proteins (FASTA file) to identify a signal peptide type: secretory protein, non-secretory protein, or transmembrane peptides. The program also checks if a protein is annotated by GO related to a secretory function.

Excel file

CBS-SignalP 4.1

Protein-protein Interactions (PPi) This program searches protein-protein interactions between proteins on the input file. Several databases are available and the program performs queries on each of these bases using the Psicquic service. The results are analysed directly on the server, and Psicquic shows interactions verified experimentally and sorts them depending on their experimental interaction detection method.

Database:
APID ChEMBL BioGrid IntAct DIP MPIDB MINT MatrixDB iRefIndex BIND Interoporc STRING Reactome-FIs Reactome InnateDB-IMEx MolCon BindingDB I2D-IMEx I2D DrugBank GeneMANIA Spike TopFind VirHostNet UniProt MBInfo mentha InnateDB bhf-ucl HPIDb BAR EBI-GOA-nonIntAct

Excel files / PPi networks

PSI-Psicquic


ProteINSIDE current Team

KASPRIC Nicolas


Workflow, web interface, database designer.

All ProteINSIDE parts has been design during PHD.


Dr. BONNET Muriel


ProteINSIDE team leader.

ProteINSIDE biological knowledge conceptor and curator.

REICHSTADT Matthieu


Helped design the tool and database.


Dr. PICARD Brigitte


ProteINSIDE conceptor biological knowledge and curator.

Former ProteINSIDE contributors


TOURNAYRE Jeremy (2013-2015): Cytoscape web interface designer for PPi. Statistics calculations for PPi (betweenness and closeness centralities).


PEROT Antoine (2015): debugging and scripting of results files (GO).