Peptide Resources

DeepQSP is a quorum-sensing peptide identification tool combining Latent Semantic Analysis (LSA), Pseudo Amino Acid Composition (PAAC), and Convolutional Neural Network (CNN) for high-precision prediction.

Deepstacked-AVPs is a novel computational model for accurate discrimination of antiviral peptides (AVPs), leveraging multi-dimensional feature fusion and stacked ensemble strategy. It employs Tri-segmentation-based Position-Specific Scoring Matrix (PSSM-TS) to capture evolutionary features, word2vec for semantic information extraction, and Composition/Transition/Distribution-Transition (CTDT) descriptors for physicochemical/structural characterization.

Diff-AMP is an integrated deep learning framework combining kinetic diffusion and attention mechanisms in reinforcement learning, automating AMP generation, identification, attribute prediction, and iterative optimization. It uses CNN and pre-training for multi-attribute prediction, with a deployed web server.

DRAMP (Data Repository of Antimicrobial Peptides) is an open-access comprehensive database containing general, patent, and clinical antimicrobial peptides (AMPs). The current version 2.0 includes 19,899 entries (2,550 newly added), comprising 5,084 general entries, 14,739 patent entries, and 76 clinical entries. Compared with APD and CAMP, it contains 14,040 non-overlapping sequences (70.56% of DRAMP), with PubMed_IDs of references included in activity information for traceability.

ENNAVIA is a deep neural network-based webserver for predicting antiviral and anti-coronavirus peptide activities, integrating deep learning and cheminformatics for efficient sequence-level screening. The model employs a hybrid architecture of convolutional neural networks (CNN) and recurrent neural networks (RNN) to automatically extract spatial features (hydrophobic patterns, charge distribution) and long-range dependencies from peptide sequences.