Related Experiment Video
Updated: Sep 29, 2025
03:37
Author Spotlight: Impact of Intergenic Interactions on Disease-Identifying Dark Biomarkers
Published on: March 1, 2024
932
MLRDFM: a multi-view Laplacian regularized DeepFM model for predicting miRNA-disease associations
Yulian Ding1, Xiujuan Lei2, Bo Liao3
1Division of Biomedical Engineering, University of Saskatchewan, 57 Campus Drive, S7N 5A9, Saskatchewan, Canada.
Briefings in Bioinformatics
|March 24, 2022
Summary
This study introduces MLRDFM, a novel model for predicting microRNA-disease associations. MLRDFM enhances deep factorization machines by incorporating similarity networks, improving prediction accuracy and reducing overfitting for disease biomarker discovery.
Area of Science:
- Bioinformatics
- Genomics
- Computational Biology
Background:
- MicroRNAs (miRNAs) are key regulators in biological processes.
- Abnormal miRNA activity is linked to various human diseases.
- Identifying disease-related miRNAs aids in biomarker discovery for disease management.
Purpose of the Study:
- To develop an advanced model for predicting novel miRNA-disease associations.
- To improve upon existing deep factorization machine (DeepFM) models.
- To enhance the accuracy and reliability of miRNA-disease association predictions.
Main Methods:
- Proposed a multi-view Laplacian regularized deep factorization machine (MLRDFM) model.
- Integrated miRNA and disease similarity networks using Laplacian regularization.
- Utilized Laplacian eigenmaps for initializing model weights.
Main Results:
- MLRDFM demonstrated improved performance and reduced overfitting compared to standard DeepFM.
- The model significantly outperformed state-of-the-art methods in miRNA-disease association prediction.
- Validation on the HMDD v3.2 dataset and case studies confirmed MLRDFM's effectiveness.
Conclusions:
- MLRDFM is a powerful tool for predicting miRNA-disease associations.
- The model's approach enhances biomarker discovery for complex diseases.
- This method offers a robust strategy for advancing genomic medicine and diagnostics.
