A powerful tool for predicting and designing antifungal peptide (AFP) using machine learning.

Welcome to Antifungipept

Antifungipept is a comprehensive tool designed for the prediction and design of antifungal peptides. It integrates advanced quantitative structure-activity relationship (QSAR) models for accurate antifungal activity prediction and provides a robust platform for rational peptide design.

Antifungipept scheme

Key Features:

  • Identification model for antifungal peptides.
  • Activity prediction models against multiple Candida species and Cryptococcus neoformans.
  • Antifungal index (AFI) for quantitatively assessing peptide effectiveness.
  • AFP20200317 database with over 400,000 putative antifungal peptide sequences.
  • Tools for sequence segmentation, mutation, and optimization guided by AFI.

Model Performance Metrics

Table 1. Results of the Antifungal Peptide Classification Model
Sample size Calibration Validation
Acc. Sen. Spec. F1 MCC Acc. Sen. Spec. F1 MCC
n=9240, 2310 a 0.95 0.95 0.95 0.95 0.90 0.89 0.90 0.89 0.89 0.79

a Sample size of calibration and validation set, respectively.

Table 2. Results of Antifungal Activity Prediction Models
Targets Calibration Validation
C. albicans (n=1266, 317) 0.69 0.90 1.23 0.66
C. krusei (n=76, 19) 0.48 0.94 1.10 0.69
C. neoformans (n=220, 55) 0.82 0.90 0.89 0.89
C. parapsilosis (n=118, 30) 0.73 0.90 1.17 0.69

Dataset Availability

Access our comprehensive datasets for training and validation:

Backend and Source Code

Antifungipept is powered by the antifungal backend. Explore and contribute to our open-source project on GitHub.

References and Citation

For detailed methodologies used in the development of Antifungipept, including the calculation of AFI and antifungal activity, please refer to the following publication:

Please cite this publication when using the data and methods from this database for academic or commercial purposes.