Next-generation antifungal peptide discovery: the synergy of artificial intelligence and omics technologies.

There is a growing concern about fungal infections and antifungal resistance among fungal species, underscoring the need for finding alternative treatments. Antifungal peptides (AFPs) are interesting and promising candidates for developing novel antifungals with high efficacy and low resistance rates. Identifying peptides with antifungal activity through classical methods is laborious and very complicated and it involves consecutive trial and error, which is expensive and time-consuming. However, novel advancements in Artificial Intelligence (AI), such as Machine Learning (ML), Deep Learning (DL), and Natural Language Processing (NLP), have brought promising success in the design and identification of novel AFPs. These achievements have improved predictions with acceptable precision and accuracy, and facilitated the peptide development with more desirable features. Since there are various limitations in AI usage in AFPs prediction models such as model complexity, limited data size, and decision processes that can affect model performance, some solutions, such as transfer learning, explainable AI (XAI), feature selection, using different important features, and genetic algorithms, can provide more valid and better performance for the prediction models. Omics technology can be a promising approach for mining genes that are responsible for producing AMPs (antimicrobial peptides) and biosynthetic gene clusters (BGCs) in natural sources of AFPs. It can be combined with ML and DL to develop novel AMPs such as AFPs. Another challenge in AFPs development is finding scalable production methods. CRISPR-Cas9, a gene-editing technique, can be utilized to enhance AFPs production in microorganisms. Besides all the advantages of AFPs as a promising treatment, improving their efficacy, ensuring safety, and appropriate delivery systems for transferring these molecules to the site of infection should be taken into account to accelerate the use of AFPs. This review highlights the role of AI in eliminating these gaps in bringing AFPs into clinical stages, and uncovers the details about preclinical and clinical studies needed to develop safe and effective AFPs with efficient delivery systems.