DefPred is a comprehensive computational platform dedicated to the identification, analysis, and design of defensins. Defensins are a major family of antimicrobial peptides (AMPs) that serve as a critical component of the innate immune system in plants, invertebrates, and vertebrates. This tool addresses the challenge of identifying novel defensins from large-scale genomic and proteomic data.
Web Server: https://webs.iiitd.edu.in/raghava/defpred/
Kaur D, Patiyal S, Arora C, Singh R, Lodhi G and Raghava GPS (2021) In-Silico Tool for Predicting, Scanning, and Designing Defensins. Front. Immunol. 12:780610. https://doi.org/10.3389/fimmu.2021.780610
This dataset is available on Zenodo at https://doi.org/10.5281/zenodo.20094615
Defensins are small, cysteine-rich cationic peptides that exhibit broad-spectrum activity against bacteria, fungi, and viruses. They are characterized by a unique structural fold stabilized by three or four disulfide bonds. DefPred provides a robust framework to classify these peptides and understand their functional properties.
- Dataset: The models were developed using a large dataset of 854 experimentally validated defensins and 854 non-defensin peptides (randomly selected from UniProt).
- Methodology: The platform utilizes several machine learning techniques, including Support Vector Machine (SVM), Random Forest, and Extra Trees, based on various sequence-derived features.
- Sequence Features: Predictions are based on amino acid composition, dipeptide composition, and physicochemical properties.
- Evolutionary Information: Uses Position-Specific Scoring Matrices (PSSM) to capture conserved evolutionary patterns unique to defensins.
- Performance: The best-performing model achieved a maximum AUROC of 0.98 and an accuracy of 94.26% on an independent dataset.
- Genome Scanning: Allows researchers to scan entire protein or genome sequences to identify potential defensin-like regions.
- In-Silico Design: A module for generating peptide analogs. Users can introduce "point mutations" to observe how changes in the sequence affect the predicted defensin score and physicochemical properties.
- Comprehensive Search: Facility to browse and search the dataset for known defensins based on source organism, peptide length, and disulfide connectivity.
- BLAST Search: Similarity search tool against the database of validated defensin sequences.
- Physicochemical Profiling: Calculates essential parameters like charge, hydrophobicity, and amphipathicity for submitted sequences.
- Antimicrobial Discovery: Identifying novel defensins as candidates for next-generation antibiotics to combat multi-drug resistant pathogens.
- Immunology Research: Studying the evolutionary conservation and functional diversity of the innate immune response across different species.
- Peptide Engineering: Optimizing defensin-based scaffolds for better stability and lower toxicity in therapeutic applications.
Prof. Gajendra P. S. Raghava (Corresponding Author)
Department of Computational Biology, Indraprastha Institute of Information Technology (IIIT Delhi), New Delhi, India.
The development of DefPred was supported by the Department of Biotechnology (DBT) and the Council of Scientific and Industrial Research (CSIR), Government of India. Infrastructure and facilities were provided by IIIT-Delhi.