Related Experiment Video
Updated: May 29, 2025
08:20
Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
Published on: October 27, 2023
1.3K
Deep learning-based prediction of autoimmune diseases.
Donghong Yang1, Xin Peng1, Senlin Zheng2
1School of Information Engineering, Jingdezhen Ceramic University, Jingdezhen, 333403, China.
Scientific Reports
|February 7, 2025
Summary
This study introduces two AI models, AutoY and LSTMY, for predicting autoimmune diseases using T-cell receptors. The AutoY model achieved high accuracy, exceeding 0.93 AUC, showing promise for early disease detection.
Area of Science:
- Immunology and computational biology.
- Focuses on the role of T cells and T-cell receptors (TCRs) in autoimmune disease pathogenesis.
Background:
- Autoimmune diseases result from immune system dysfunction, with complex etiologies involving genetics and environmental factors.
- T cells are crucial in immune responses and disease development, with TCRs implicated in various autoimmune conditions.
- Accurate prediction and early detection of autoimmune diseases remain significant challenges.
Purpose of the Study:
- To develop and evaluate computational models for predicting autoimmune diseases based on T-cell data.
- To assess the efficacy of deep learning models, specifically convolutional neural networks (CNNs) and Long Short-Term Memory (LSTM) networks, for this predictive task.
Main Methods:
- Proposed two predictive models: AutoY (CNN-based) and LSTMY (bidirectional LSTM with attention mechanism).
- Utilized T-cell receptor (TCR) data as input for disease prediction.
- Evaluated model performance using metrics such as the Area Under the ROC Curve (AUC).
Main Results:
- Both AutoY and LSTMY models demonstrated strong performance in predicting four types of autoimmune diseases.
- The AutoY model slightly outperformed LSTMY, with an average AUC exceeding 0.93 across all tested diseases.
- Exceptional AUC values of 0.99 were achieved for type 1 diabetes and multiple sclerosis predictions.
- The models exhibited high accuracy, stability, and generalization capabilities.
Conclusions:
- The developed AI models show significant potential as tools for the accurate and non-invasive prediction of autoimmune diseases.
- TCR bank data can be leveraged for early detection, offering a promising avenue for clinical application.
- Further validation and application of these models could enhance autoimmune disease management and treatment strategies.
