VirFacPred is a computational web server developed for predicting virulent and non-virulent proteins from primary protein sequences.
The tool uses a wide range of information and computational techniques, including machine learning, BLAST-based similarity search, and MERCI-based motif scanning. VirFacPred is useful for identifying proteins that may contribute to pathogenicity, host invasion, survival, immune evasion, and virulence mechanisms.
Web Server: https://webs.iiitd.edu.in/raghava/virfacpred/
This dataset is also available on Zenodo at
Virulent proteins are proteins that help pathogens cause disease in a host. These proteins may help pathogens attach to host cells, invade tissues, escape immune responses, survive inside the host, or damage host systems.
Identification of virulent proteins is important for understanding pathogenic mechanisms and for discovering potential drug targets, vaccine candidates, and diagnostic markers.
VirFacPred was developed to predict virulent proteins using sequence-based computational approaches.
Data Compilation: The models were trained on a large dataset containing 8233 virulent proteins and 8233 non-virulent proteins.
Methodology: VirFacPred uses machine learning-based prediction, BLAST similarity search, and MERCI motif scanning. The web server provides both individual and hybrid prediction approaches.
Predictive Modeling: Allows users to submit protein sequences and predict whether they are virulent or non-virulent.
Batch Prediction: Users can submit multiple protein sequences in FASTA format.
Machine Learning Model: The server provides an amino acid composition-based Random Forest model.
Hybrid Model: The best prediction option combines Random Forest, BLAST, and MERCI.
Threshold Selection: Users can select different threshold values for prediction.
Email Option: Users can optionally provide an email address to receive prediction results.
Protein Designing: The design module allows users to submit one or multiple protein sequences in FASTA format.
Single-Mutation Analysis: The server generates all possible single-mutant sequences.
Mutation-Based Prediction: The generated mutants are further predicted using the machine learning model.
Application: This module can help identify mutations that may reduce virulence potential.
MERCI-Based Motif Scanning: VirFacPred uses MERCI to scan query protein sequences for virulent motifs.
Motif Mapping: The motif scan module helps identify important motifs present in a query protein sequence.
Batch Motif Search: Users can submit multiple protein sequences in FASTA format.
Purpose: This module helps users understand whether a protein contains sequence motifs associated with virulent or non-virulent proteins.
Similarity-Based Search: The BLAST module searches query protein sequences against a database of known virulent and non-virulent proteins.
Prediction Rule: If the query sequence gets a hit in the virulent/non-virulent database, the server uses similarity information for prediction.
Multiple Sequence Submission: Users can submit multiple protein sequences in FASTA format.
E-value Selection: Users can choose different BLAST E-value cutoffs.
VirFacPred provides several useful links and modules:
- Prediction
- Design
- Motif Scan
- BLAST Search
- Download
- Algorithm
- Help Page
- Team
- Contact
The download section provides access to the standalone version and datasets.
Virulence Factor Prediction: VirFacPred can help identify proteins that may contribute to microbial virulence.
Pathogen Biology: The tool can support studies on how pathogens invade, survive, and cause disease in host organisms.
Drug Target Discovery: Predicted virulent proteins may be prioritized as potential antimicrobial drug targets.
Vaccine Candidate Screening: Virulent proteins can be useful for identifying possible vaccine candidates.
Comparative Genomics: VirFacPred can support proteome-level screening of pathogenic organisms.
Functional Annotation: The tool can help annotate hypothetical or uncharacterized proteins as virulent or non-virulent.
Prof. Gajendra P. S. Raghava
Head, Department of Computational Biology
Indraprastha Institute of Information Technology Delhi
Okhla Phase III
New Delhi-110020, India
Email: raghava@iiitd.ac.in, raghavagps@gmail.com
Support and funding details were not specified on the accessed VirFacPred web pages.
The server is hosted by the Department of Computational Biology, Indraprastha Institute of Information Technology Delhi.