Predicting miRNA-disease association based on inductive matrix completion.

Area of Science:

• Genomics and Bioinformatics
• Computational Biology
• Molecular Medicine

Background:

• MicroRNAs (miRNAs) are crucial regulators in biological processes and human diseases.
• Experimental identification of miRNA-disease associations is costly and complex.
• Computational approaches offer an efficient complement for predicting these associations.

Purpose of the Study:

• To develop and validate a novel computational model, Inductive Matrix Completion for MiRNA-Disease Association (IMCMDA), for predicting miRNA-disease associations.
• To integrate multiple similarity measures for enhanced prediction accuracy.
• To assess the model's performance and robustness across various human diseases.

Main Methods:

• Developed the IMCMDA model utilizing Inductive Matrix Completion.
• Integrated miRNA functional similarity, disease semantic similarity, and Gaussian interaction profile kernel similarity.
• Calculated integrated miRNA and disease similarities to infer missing associations.

Main Results:

• Achieved an Area Under the Curve (AUC) of 0.8034 via leave-one-out cross-validation, outperforming previous models.
• Successfully predicted known miRNA-disease associations for Colon Neoplasms, Kidney Neoplasms, and Lymphoma (42-45/50 confirmed).
• Demonstrated high accuracy in predicting novel associations for Breast Neoplasms (50/50 confirmed) and Esophageal Neoplasms (49/50 confirmed).

Conclusions:

• The IMCMDA model provides an effective computational tool for predicting miRNA-disease associations.
• The model demonstrates high accuracy and robustness, validated across multiple disease datasets.
• IMCMDA facilitates the discovery of novel miRNA-disease links, aiding biomedical research.