Related Experiment Video
Updated: Oct 3, 2025
06:45
Quantifying the Antifungal Activity of Peptides Against Candida albicans
Published on: January 13, 2023
2.4K
Accelerating the discovery of antifungal peptides using deep temporal convolutional networks
Vishakha Singh1, Sameer Shrivastava2, Sanjay Kumar Singh1
1Department of Computer Science and Engineering, Indian Institute of Technology (BHU), Varanasi, 221005, Uttar Pradesh, India.
Briefings in Bioinformatics
|February 13, 2022
Summary
Researchers developed a machine learning model to rapidly discover antifungal peptides (AFPs) for drug discovery. This AI tool accelerates the identification of novel antifungal molecules, aiding in the fight against fungal infections.
Area of Science:
- Bioinformatics
- Computational Biology
- Drug Discovery
Background:
- Machine intelligence accelerates drug discovery, with AI interest surging due to COVID-19.
- Mucormycosis is a significant post-COVID-19 fungal complication, particularly in immunocompromised individuals.
- Traditional methods for identifying antifungal peptides (AFPs) are laborious and hindered by limited datasets.
Purpose of the Study:
- To develop an automated approach for discovering novel antifungal molecules (AFPs).
- To accelerate the development of antifungal medications by identifying plant and animal-derived AFPs.
- To address the challenge of limited AFP datasets using transfer learning.
Main Methods:
- Proposed a temporal convolutional network (TCN)-based binary classification model.
- Employed transfer learning, pre-training the model on antibacterial peptides.
- Validated model performance using statistical tests (Kruskal-Wallis H, Tukey HSD).
Main Results:
- Achieved 94% accuracy and precision in predicting AFPs.
- The TCN classifier significantly outperformed existing state-of-the-art models.
- Identified potent AFPs in animal (Histatin) and plant (Snakin) proteins.
- Developed a publicly accessible web application for AFP identification.
Conclusions:
- The developed TCN model offers a robust and efficient method for antifungal peptide discovery.
- This AI-driven approach can accelerate the development of new antifungal therapies.
- The freely available web app facilitates broader research and application in identifying AFPs.
