Related Experiment Video
Updated: Jun 17, 2025
09:47
Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches
Published on: December 15, 2023
1.0K
A deep learning drug screening framework for integrating local-global characteristics: A novel attempt for limited
Ying Wang1, Yangguang Su1, Kairui Zhao1
1Department of Pharmacogenomics, College of Bioinformatics Science and Technology, Harbin Medical University, Harbin, Heilongjiang, 150081, China.
Heliyon
|August 12, 2024
Summary
A new convolutional neural network (CNN) model, LGCNN, rapidly screens drugs using limited data by integrating molecular and interaction features. This approach aids therapeutic drug discovery during outbreaks like COVID-19.
Area of Science:
- Computational biology
- Drug discovery
- Machine learning
Background:
- Drug discovery faces challenges with insufficient data during outbreaks.
- Limited data hinders traditional drug screening models.
Purpose of the Study:
- To develop a novel drug screening model, LGCNN, for rapid drug discovery with limited data.
- To integrate local and global features of drug molecular structures and drug-target interactions.
Main Methods:
- Proposed LGCNN, a convolutional neural network (CNN) model.
- Integrated local and global features for drug screening.
- Applied LGCNN to anti-SARS-CoV-2 drug screening for COVID-19 therapeutics.
Main Results:
- LGCNN demonstrated superior performance over state-of-the-art methods with limited data.
- LGCNN successfully identified potential therapeutics for COVID-19.
- The model showed advantages in predicting multi-target drug-target interactions.
Conclusions:
- LGCNN offers a novel approach for rapid drug screening in data-scarce emergency situations.
- The model's cross-coronavirus generalizability suggests broader applications.
- LGCNN enhances drug discovery efficiency by maximizing information from limited datasets.
