DRAMP 5.0: an update integrating sequence and structure in a novel FASTS format for AI-driven design of antimicrobial peptides.

Antimicrobial peptides (AMPs) represent promising candidates for combating multidrug-resistant bacterial infections; however, their clinical translation has been hampered by inherent limitations, including poor proteolytic and serum stability, unfavorable pharmacokinetic profiles, and cytotoxicity or hemolytic effects. In recent years, artificial intelligence (AI)-driven peptide design has emerged as a transformative strategy to accelerate AMP discovery and optimization by navigating vast sequence spaces and predicting multi-objective properties. Nevertheless, existing AMP databases predominantly catalog sequence and activity information, with insufficient annotation of non-natural amino acid and chemical modifications. In this update, the Data Repository of Antimicrobial Peptides (DRAMP) introduces a novel FASTS format, which incorporates SMILES representations into FASTA annotation lines, enabling the simultaneous storage of sequence and structural information within a single text file. This unified format allows accurate representation of peptides containing non-natural residues and chemical modifications, facilitates structure-aware sequence analysis, and supports AI/ML-based peptide prediction and design. DRAMP 5.0 currently contains 32,296 entries (2,036 newly added). In addition, 22,298 peptide sequences have been annotated in the newly proposed FASTS format. The updated DRAMP provides an enhanced resource for AMP research, cheminformatics, and AI-driven peptide discovery, and remains freely available at https://dramp.cpu-bioinfor.org/.