Peptide Resources

ConoServer specializes in sequences and structures of conopeptides, toxins secreted by marine cone snails. As carnivorous gastropods, cone snails use toxin cocktails to disrupt prey nervous systems, with toxins targeting neural receptors, channels, and transporters—valuable for physiological research and drug development. Since 2008, entries have nearly doubled, with updated annotations (species descriptions, activity data, sequence identification) and new tools for transcriptomic/proteomic analysis (precursor sequence standardization, MS-based toxin identification). Automatically updated statistics on classification, 3D structures, toxin-bearing species, and ER signal sequence conservation offer systematic insights.

CpACpP is a predictor for cell-penetrating anticancer peptides (Cp-ACPs) using RF, SVM, XGBoost algorithms with CPP/ACP subpredictors and multilayered RFE feature selection.

C-PAmP is a professional database for computationally predicted plant antimicrobial peptides, containing 15,174,905 peptides (5-100 amino acids) derived from 33,877 proteins of 2,112 plant species in UniProtKB/Swiss-Prot. Using an improved pseudo amino acid concept classification algorithm (10-fold cross-validation accuracy 0.91, sensitivity 0.93, specificity 0.90), it supports retrieval by peptide/protein sequence, protein accession number, and species, allowing viewing of prediction probability scores, CAMP classification, PhytAMP ID (where applicable), and visualization of high-concentration antimicrobial peptide regions in proteins. It provides targeted data support for experimental screening and drug design of plant-derived antimicrobial peptides.

CPPred-FL is a sequence-based predictor for large-scale identification of cell-penetrating peptides, leveraging feature representation learning from 45 well-trained random forest models with integration of class/probabilistic information and feature space optimization.

CPPred-RF is a sequence-based predictor using random forest algorithm, integrating multi-feature descriptors and feature selection for two-layer prediction of cell-penetrating peptides and their uptake efficiency.