iAMP-2L

iAMP-2L predicts whether peptide sequences are antimicrobial and assigns identified AMPs to functional categories to support AMP annotation and functional assignment.

Key Features:

• Two-level prediction: A first level discriminates AMPs from non-AMPs and a second level assigns one or more functional types to peptides identified as AMPs.
• Multi-label classification: Assigns AMPs to ten functional categories: Antibacterial, Anticancer/tumor, Antifungal, Anti-HIV, Antiviral, Antiparasital, Anti-protist, Chemotactic activity, Insecticidal, and Spermicidal.
• Pseudo Amino Acid Composition (PseAAC): Encodes sequence-order information and five physicochemical properties as input features.
• Fuzzy K-Nearest Neighbor (FKNN) algorithm: Employs a fuzzy nearest-neighbor classifier to accommodate ambiguous or multi-type assignments.
• Sequence-based input: Operates solely on peptide primary sequences without requiring structural or experimental data.

Scientific Applications:

• Basic research: Enables functional annotation and comparative analysis of AMPs for studies of distribution, evolution, and biological roles.
• Drug development: Supports identification and prioritization of candidate antimicrobial peptides for therapeutic discovery against pathogens and diseases.

Methodology:

Two-level multi-label prediction using PseAAC feature encoding (including five physicochemical properties and sequence-order information) combined with a Fuzzy K-Nearest Neighbor (FKNN) classifier.