MLAFP-XN: Leveraging neural network model for development of antifungal peptide identification tool

Abstract

Infectious fungi have been an increasing global concern in the present era. A promising approach to tackle this pressing concern involves utilizing Antifungal peptides (AFP) to develop an antifungal drug that can selectively eliminate fungal pathogens from a host with minimal toxicity to the host. Accordingly, identifying precise therapeutic antifungal peptides is crucial for developing effective drugs and treatments. This study proposed MLAFP-XN, a neural network-based strategy for accurately detecting active AFP in sequencing data to achieve this objective. In this work, eight feature extraction techniques and the XGB feature selection strategy are utilized together to present an enhanced methodology. A total of 24 classification models were evaluated, and the most effective four have been selected. Each of these models demonstrated superior accuracy on independent test sets, with respective scores of 97.93 %, 99.47 %, and 99.48 %. Our model outperforms current state of the art methods. In addition, we created a companion website to demonstrate our AFP recognition process and use SHAP to identify the most influential properties.

Keywords: Antifungal drug; Antifungal peptide; Drug discovery; Feature extraction; Feature selection; Neural network.

Fig. 1

The overall workflow of the study: Data collection, feature encoding, feature selection process, applied evaluation method, development of the MLAFP-XN model, and deployment based on the proposed model.

Fig. 2

Feature selection process of the study. Features are selected based on feature importance scores and the selected features are combined accordingly. The XGB feature selection approach was applied to the merged dimension and delivered the 500 optimal features.

Fig. 3

MLAFP-XN model's architecture, input layer, an output layer, and four hidden layers with ReLU and sigmoid activation function for 500-dimensional features.

Fig. 4

Receiver operating characteristics curves (ROC) and precision-recall curves (PR) analysis of all models with AUC scores, where the blue straight line is the random line. (A) Antifp_DS1 dataset's ROC curves (B) Antifp_DS1 dataset's PR curves, (C) Antifp_DS2 dataset's ROC curves (D) Antifp_DS2 dataset's PR curves, (E) Antifp_Main dataset's ROC curves (F) Antifp_Main dataset's PR curves. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 5

SHAP feature analysis of ML-AFP-XN model, (A) Antifp_DS1 dataset's outcome, (B) Antifp_DS2 dataset's results, (C) Antifp_Main dataset's findings.