AI/ML-driven Fung-AI pipeline generates novel antifungal peptides active against *Fusarium graminearum* and *Candida albicans*

Emerging fungal pathogens pose a significant threat to global health and food security, with rising antimicrobial resistance limiting treatment options. Traditional drug discovery is slow and costly, failing to keep pace with evolving resistance. Antimicrobial peptides (AMPs) offer a promising alternative due to their broad-spectrum activity and novel mechanisms. This study addresses the urgent need for new antifungals by leveraging artificial intelligence to accelerate the discovery of de novo peptide candidates.

Researchers developed the Fung-AI pipeline, an AI/ML-driven approach for antifungal discovery. A generative adversarial network (GAN) was trained to create novel peptide sequences. In silico antifungal and hemolytic classifiers then prioritized these AI-generated peptides. From approximately 10,000 candidates, thirteen peptides were selected for experimental validation. These were tested for antifungal activity against the wheat pathogen Fusarium graminearum and the human pathogen Candida albicans using MIC assays, and against C. auris. Cytotoxicity was assessed in HepG2 human liver carcinoma cells.

The Fung-AI pipeline successfully identified novel antifungal peptides. > Five of the thirteen peptides displayed mild antifungal activity against Fusarium graminearum, with minimal inhibitory concentrations (MICs) ranging from 250 µg/mL to 500 µg/mL. Four of these five peptides also showed activity against Candida albicans, with an MIC of 500 µg/mL. Two of the AI-generated antifungal peptides demonstrated low cytotoxicity in HepG2 cells, with LC50 values of >704.2 µg/mL, suggesting potential as scaffolds for therapeutic optimization. However, none of the peptides considerably inhibited the emerging pathogen C. auris, indicating the necessity for pathogen-specific candidate selection and optimization.