Antifungal peptides (AFPs) have emerged as promising alternatives to conventional antifungal agents due to their broad-spectrum activity, low toxicity, and reduced propensity for resistance development. However, it is still a challenging task to quickly identify possible AFPs from large protein databases. The current antifungal therapies and medications are widely acknowledged as insufficient due to their associated adverse effects. To assess the effectiveness of AFPs in the human system, developing a reliable, intelligent model becomes imperative for the meticulous and precise identification of these peptides endowed with antifungal properties. Therefore, developing a machine learning framework is imperative to identify AFP effectively. In this paper, we present a novel approach for the identification of antifungal peptides using the Extra Tree Classifier with fusion features, including amino acid composition (AAC), dipeptide composition (DPC), and pseudo amino acid composition (PseAAC). Combining these hybrid features enhances classification accuracy and facilitates the efficient screening of potential AFP candidates. Through the implementation of a five-fold cross-validation strategy, the obtained results demonstrated the outstanding performance of our model, achieving an accuracy rate of 91.29% and an area under the curve (AUC) value of 0.96 in the identification of antifungal peptides on the training dataset. Our proposed model, named iAFP-ET, demonstrated exceptional performance, surpassing existing computational models and attaining the highest level of accuracy. The development of this model is expected to have a significant impact on research in academia, with an important contribution towards the growth of Proteomics and drug development.

Keywords: Antifungal peptides, Extra Tree Classifier, fusion features, Amino acid composition, Dipeptide composition, Pseudo amino acid composition.