Related Experiment Video
Updated: Jul 7, 2025
09:40
Characterization of Functionally Associated miRNAs in Glioblastoma and their Engineering into Artificial Clusters for Gene Therapy
Published on: October 4, 2019
5.7K
A vector projection similarity-based method for miRNA-disease association prediction
Guobo Xie1, Weijie Xie1, Guosheng Gu1
1School of Computer, Guangdong University of Technology, Guangzhou, 510000, China.
Analytical Biochemistry
|December 20, 2023
Summary
This study introduces a novel Vector Projection Similarity-based Method for miRNA-disease Association Prediction (VPSMDA). VPSMDA improves upon existing methods by better accounting for known and unknown miRNA-disease associations, achieving superior prediction accuracy.
Area of Science:
- Bioinformatics
- Computational Biology
- Genomics
Background:
- Gaussian interaction profile kernel similarity (GIPS) is commonly used in miRNA-disease association prediction.
- GIPS overlooks the specificity of the miRNA-disease association matrix, particularly zero values representing undiscovered relationships.
Purpose of the Study:
- To propose a novel method, VPSMDA, for miRNA-disease association prediction.
- To enhance similarity measures by incorporating both known and unknown miRNA-disease associations.
- To improve the accuracy of predicting potential miRNA-disease links.
Main Methods:
- Developed a Vector Projection Similarity-based Method for miRNA-disease Association Prediction (VPSMDA).
- Introduced three projection rules and logistic functions to the miRNA-disease association matrix.
- Integrated vector projection similarity with the original matrix to create an improved similarity matrix.
- Constructed a weighted matrix using neighbor information to reduce noise.
Main Results:
- VPSMDA demonstrated superior performance compared to seven state-of-the-art methods in LOOCV and 5-fold CV experiments, achieving higher AUC values.
- A case study showed VPSMDA successfully predicted top 10 associations for three human diseases, with 10, 9, and 10 predictions confirmed by recent biomedical resources.
Conclusions:
- VPSMDA offers a more effective approach to miRNA-disease association prediction by considering matrix specificity.
- The method shows significant potential for identifying novel miRNA-disease relationships.
- VPSMDA outperforms existing prediction models and provides reliable predictions validated by external data.
