Peptide Resources

MBPDB is a comprehensive database of bioactive peptides derived from mammalian milk proteins, systematically compiling peptides with antimicrobial, ACE-inhibitory, antioxidant, etc., functions. Sourced from systematic literature mining across hundreds of studies, it supports searches by function, peptide sequence, or protein origin, and visually maps active peptides to parent protein sequences for intuitive high-abundance site identification.

Meta-iAVP is the first meta-learning-based predictor for antiviral peptides (AVPs), achieving precise identification by integrating prediction scores from multiple machine learning algorithms as feature representations. Innovatively using outputs of base classifiers (e.g., Random Forest) as meta-features, combined with multi-dimensional sequence features (amino acid composition, physicochemical properties), it optimizes prediction via a two-layer ensemble framework.

MFPPDB is a database integrating multi-functional plant-derived therapeutic peptides, containing 1,482,409 single/multi-functional peptides from 121 fundamental plant species, covering 41 functional categories such as antibacterial, antifungal, anti-HIV, antiviral, and anticancer. The database provides physicochemical properties (isoelectric point, molecular weight, sequence), plant sources, and functional prediction information of peptides. Through validation by matching with 9 authoritative functional peptide databases, at least 293,408 peptides are confirmed to have functional potential.

miPepBase is a database focusing on experimentally validated molecular mimicry peptides, containing sequences of mimic proteins and peptides from hosts (and model organisms) and pathogens, covering autoimmune disease associations, protein physicochemical properties (amino acid composition, isoelectric point, molecular weight), and taxonomic data. It supports retrieval by autoimmune disease types, host-pathogen taxonomic combinations, and keywords, integrating BLAST sequence alignment tools to provide data support for analyzing autoimmune mechanisms triggered by molecular mimicry and epitope research.

MLACP is a webserver for predicting anticancer peptides (ACPs) based on support vector machine (SVM) and random forest (RF), using sequence features like amino acid composition, dipeptide composition, atomic composition, and physicochemical properties. The RF method achieves 88.7% accuracy and 0.78 MCC on benchmark datasets, outperforming existing methods, with a public service available.