VirHostnet 3.0-beta release

VirHostNet 3.0: towards systems biology of virus/host interactions.

VirHostNet is a bioinformatic information system dedidacted to the biocuration, data integration, reproducible systems-level analysis and visualisation of Virus / Host protein-protein interactions Network based on graph theory. VirHostNet is an open and gold standard knowledgebase shared in PSI MITAB 2.5 format using the PSICQUIC webservice and distributed through the NDEx platform. VirHostNet is FAIR and is recognized as a COVID-19 ressource by Elixir bio.tools, the European Virus Bioinformatics Center and FAIRsharing.org (doi:10.25504/fairsharing.m3316).

About March 2022 Release ?

Quick search

Quick search by using UniprotKb protein accession number and primary name, keyword annotation, NCBI taxonomy (taxon name/species name) or publication PMID and directly visualise their virus/host protein-protein interaction network based on the Cytoscape.js graph theory javascript library.

Advanced preferences

VirHostNet 2.0: surfing on the web of virus/host molecular interactions data.
Thibaut Guirimand, Stéphane Delmotte, Vincent Navratil
Nucleic Acids Research, Volume 43, Issue D1, 28 January 2015, Pages D583–D587
https://doi.org/10.1093/nar/gku1121

here

Blast protein sequence search

Protein sequence(s) search using blastp (Altschul et al. 1990) against a VirHostNet interacting protein sequences database (viruses + host sequences).

Browse by Viruses taxonomy.

Browse Virus/Host (VH) protein-protein interaction dataset according to taxonomy lineage of viruses (Baltimore/Family/Species/Leaf taxon).

Browse by Keywords annotations

keyword

For each category, the number of annotated proteins (#Annotated proteins), protein interacting with those proteins (#Interacting proteins) and the number of virus/host, virus/virus protein-protein (#Virus/host interactions) are provided.

Host cell receptor for virus entry [KW-1183] Virus/Host network
Innate immunity [KW-0399] Virus/Host network
Autophagy [KW-0072] Virus/Host network

Interology web service

Interology

.fasta

Network summary.

VirHostNet model virus/host interactome as an undirected graph both for visualisation and data exploration. Indeed, graph theory provides some metrics which are helpfull to rank host and viral proteins according their centrality and their location within the whole set of possible interactions identified experimentally. The viral degree of a protein (kv) represents the number of viral partners of a protein.The host degree of a protein (kh) represents the number of host partners of a protein.
The Figures below are drawn directly from the virhostnet backend with the d3 data-driven document, javascript library.

Fig. 1. Histogram distribution of the viral degree (kv) of host proteins computed within the whole virus/host protein-protein interaction network.

In this release of VirHostNet, a total of host proteins are annotated to physically interact with at least one viral proteins (kv >=1) and might be considered as viral host factors. A total of and host proteins were allready found to interact experimentally with at least 2 (kv >=2) and 10 (kv >=10) viral proteins respectively.

Fig. 2. Scatterplot between the viral (kv) and host degree (kh) of host factor proteins.

Human data

Fig. 3. Histogram distribution of the viral degree (kv) of human proteins computed within the whole virus/human protein-protein interaction network.

In this release of VirHostNet, a total of human proteins are annotated to physically interact with at least one viral proteins (kv >=1) and might be considered as viral host factors. A total of and human proteins were allready found to interact experimentally with at least 2 (kv >=2) and 100 (kv >=100) viral proteins respectively.

Expert biocuration of virus/host protein-protein interactions

Since 2006, the VirHostNet team is involved in the manual biocuration of virus/host and virus/virus protein-protein interactions by following international standard (MITAB 2.5 format). VirHostNet use a fast biocuration process (Accession Number of proteins, Experimental methods, PMID) and is complementary to more in depth biocuration actions including those promoted by the IMEX consortium (MITAB 2.7 format).
The Figure below is drawn with the rickshaw javascript toolkit.

Biocuration

Year	PMID	Title	Authors	Journal	#vp	#hp	#ppi	#ppi(nr)	PSI-MI
Year	PMID	Title	Authors	Journal	#vp	#hp	#ppi	#ppi(nr)	PSI-MI

SARS-CoV-2 release: real-time biocuration, reproducible and FAIR share systems biology research against Covid-19.

To better respond to the Covid-19 crisis, comprehensive manual biocuration of Coronaviridae / host protein-protein interactions was quickly achieved by our team in early 2020. This includes the experimental interactomes generated by the Krogan Lab Gordon DE et al. Science. 2020 Dec (Download in PSI-MITAB 2.5 - or visualise ) and Gordon DE et al. Nature. 2020 Jul (Download in PSI-MITAB2.5 or visualise ).

This release include 55.115 virus-host and virus-virus manually biocurated protein-protein interactions, including more than 15.000 new interactions since the last March 2021 release.
The use of the Cytoscape.js library for the visualisation of virus/host network.
A VirHostNet SOLR PSICQUIC webservice that provides openaccess annotations in PSI MI-TAB 2.5 format (download)
The use of UniProtkb/Swiss-Prot TrEMBL (release January 2022) and ReFSeQ (release 211 March 2022) as well as GenBank as references (March 2022)
The use of PSICQUIC webservice (March 7th 2022) to download INTACT, MINT, BIOGRID, DIP, Uniprot, MatrixDB, HPIDb, IMEx protein-protein interaction data in PSI-MITAB 2.5 format
Virus' host range annotations provided by Viralzone, Virus-Host DB, and in-house text-mining of genbank are now integrated and used to highlight biologically relevant virus/host ppi
The use of UniProtkb chain feature key - including mature peptides - to annotate protein-protein interactions at a finer scale
The use of MIscore scoring function to measure the quality of each non-redundant interaction based on independant experiments information
Network-based nodes and edges properties (degree, betweeness, ...)
A BLAST quick search and webservice to infer novel virus/host ppi from sequences based on the assumption of interology.