Related Experiment Video
Updated: Jun 21, 2025
03:37
Author Spotlight: Impact of Intergenic Interactions on Disease-Identifying Dark Biomarkers
Published on: March 1, 2024
685
Kernel Bayesian logistic tensor decomposition with automatic rank determination for predicting multiple types of
1School of Mathematics and Statistics, Xiamen University of Technology, Xiamen, China.
Plos Computational Biology
|July 8, 2024
Summary
This study introduces KBLTDARD, a novel computational framework for identifying microRNA (miRNA) and disease associations. KBLTDARD enhances prediction accuracy by integrating biological networks and employing advanced Bayesian tensor decomposition methods.
Area of Science:
- Bioinformatics
- Computational Biology
- Genomics
Background:
- Identifying microRNA (miRNA)-disease associations is vital for understanding disease mechanisms.
- Computational models offer efficiency and scalability over traditional experiments but often struggle with hyperparameter tuning and complex nonlinear relationships.
- Existing tensor decomposition models require manual hyperparameter specification, limiting efficiency and generalization.
Purpose of the Study:
- To propose a novel framework, KBLTDARD, for identifying multiple types of miRNA-disease associations.
- To overcome limitations of existing models by enabling automatic hyperparameter search and incorporating auxiliary information.
- To improve the accuracy and efficiency of predicting miRNA-disease associations.
Main Methods:
- KBLTDARD integrates information from biological networks and high-order association networks to derive precise miRNA and disease similarities.
- A combination of logistic tensor decomposition and Bayesian methods with sparse-induced priors facilitates automatic hyperparameter optimization.
- An efficient deterministic Bayesian inference algorithm is employed for computational efficiency.
Main Results:
- KBLTDARD demonstrated superior performance in predicting new types of miRNA-disease associations, achieving higher Top-1 precision, recall, and F1 scores.
- The framework showed improved prediction accuracy for new miRNA-disease triplets, evidenced by higher AUPR, AUC, and F1 values compared to state-of-the-art methods.
- Case studies confirmed the practical efficiency of KBLTDARD in predicting diverse miRNA-disease associations.
Conclusions:
- KBLTDARD offers a robust and efficient computational approach for identifying multiple types of miRNA-disease associations.
- The framework's ability to automatically handle hyperparameters and incorporate auxiliary data enhances its predictive power and generalization.
- KBLTDARD represents a significant advancement in computational methods for miRNA-disease association studies.
