Related Experiment Video
Updated: Aug 28, 2025
Characterization of Functionally Associated miRNAs in Glioblastoma and their Engineering into Artificial Clusters for Gene Therapy
Published on: October 4, 2019
Inferring human miRNA-disease associations via multiple kernel fusion on GCNII
Shanghui Lu1,2, Yong Liang1,3, Le Li1
1School of Computer Science and Engineering, Macau University of Science and Technology, Taipa, China.
Abstract:
Increasing evidence shows that the occurrence of human complex diseases is closely related to the mutation and abnormal expression of microRNAs(miRNAs). MiRNAs have complex and fine regulatory mechanisms, which makes it a promising target for drug discovery and disease diagnosis. Therefore, predicting the potential miRNA-disease associations has practical significance. In this paper, we proposed an miRNA-disease association predicting method based on multiple kernel fusion on Graph Convolutional Network via Initial residual and Identity mapping (GCNII), called MKFGCNII. Firstly, we built a heterogeneous network of miRNAs and diseases to extract multi-layer features via GCNII. Secondly, multiple kernel fusion method was applied to weight fusion of embeddings at each layer. Finally, Dual Laplacian Regularized Least Squares was used to predict new miRNA-disease associations by the combined kernel in miRNA and disease spaces. Compared with the other methods, MKFGCNII obtained the highest AUC value of 0.9631. Code is available at https://github.com/cuntjx/bioInfo.
Insights
Predicting microRNA-disease associations is crucial for drug discovery. A new method, MKFGCNII, uses graph convolutional networks and kernel fusion to accurately identify these links, achieving a high AUC of 0.9631.
Area of Science:
- Biomedical Informatics
- Computational Biology
- Genomics
Background:
- Complex human diseases are linked to microRNA (miRNA) mutations and abnormal expression.
- miRNAs regulate biological processes, making them key targets for disease diagnosis and drug development.
- Accurate prediction of miRNA-disease associations is vital for advancing personalized medicine.
Purpose of the Study:
- To develop a novel computational method for predicting potential miRNA-disease associations.
- To leverage advanced deep learning techniques for enhanced feature extraction and association prediction.
- To provide a robust tool for identifying novel therapeutic targets and diagnostic biomarkers.
Main Methods:
- Proposed MKFGCNII, a method employing multiple kernel fusion on a Graph Convolutional Network with Initial residual and Identity mapping (GCNII).
- Constructed a heterogeneous network of miRNAs and diseases to extract multi-layer features using GCNII.
- Applied multiple kernel fusion to integrate embeddings from different layers and utilized Dual Laplacian Regularized Least Squares for prediction.
Main Results:
- The MKFGCNII method achieved a high Area Under the Curve (AUC) value of 0.9631 in predicting miRNA-disease associations.
- Demonstrated superior performance compared to existing methods in identifying potential miRNA-disease links.
- The developed heterogeneous network and fusion strategy effectively captured complex biological relationships.
Conclusions:
- MKFGCNII offers a powerful and accurate approach for predicting miRNA-disease associations.
- This method holds significant potential for accelerating drug discovery and improving disease diagnosis.
- The findings underscore the importance of integrating network-based approaches with deep learning for biological data analysis.
Related Concept Videos
01:22MicroRNAs
01:11Genome-wide Association Studies-GWAS
GWAS does not require the identification of the target gene involved in...
02:34DNA Microarrays