Abstract
Wound healing is a complex process that can be compromised by bacterial infections, leading to delayed healing and an increased risk of complications. The aim of this study was to design and develop a novel antibacterial peptide, Healitide-GP1, which could be used in the treatment of infected wounds. A machine learning-based approach was used to identify the key features of WHPs and a genetic algorithm was used to generate new peptide sequences. Healitide-GP1 was synthesized and functionally validated. It showed high cytocompatibility (> 200 µg/mL in human dermal fibroblasts (HDF) and human keratinocytes (HaCaT)), improved wound closure (48% and 52% after 24 h, respectively) and strong antibacterial activity against Staphylococcus aureus (MIC: 12.5 µg/mL) and Escherichia coli (MIC: 25 µg/mL). Bioinformatics analyses revealed the unique hydrophobic motif and distinct evolutionary positioning of Healitide-GP1, suggesting a novel mechanism of action. These results emphasize the potential of Healitide-GP1 as a promising therapeutic candidate for the treatment of infected wounds. Our study demonstrates the successful integration of machine learning, bioinformatics and experimental validation in the development of therapeutic peptides and provides a valuable framework for the discovery of novel treatments for wound infections.