A rapid prediction model for MiRNA-LncRNA interactions utilizing multi-view projection fusion and random parallel...

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• MicroRNA (miRNA) and long non-coding RNA (lncRNA) interactions are key regulators of gene expression.
• Dysregulated miRNA-lncRNA interactions are implicated in various diseases.
• Efficient computational methods are needed for predicting these interactions due to increasing data complexity.

Purpose of the Study:

• To develop a rapid and accurate computational model for predicting miRNA-lncRNA interactions.
• To address limitations of existing models, including single similarity metrics and computational inefficiency.
• To explore the utility of similarity-based network topologies in interaction prediction.

Main Methods:

• Constructed a comprehensive miRNA/lncRNA similarity profile.
• Developed a Multi-view Projection Fusion technique to update the interaction matrix using diverse network topologies.
• Proposed a Divide-Parallel-Combine (DPC) framework for parallel acceleration via truncated matrix factorization.

Main Results:

• The MVPFDPC model demonstrated high accuracy in predicting miRNA-lncRNA interactions.
• The DPC framework significantly improved computational speed for large-scale matrix tasks.
• Experiments on three datasets validated the model's performance.

Conclusions:

• MVPFDPC provides an efficient and accurate solution for predicting miRNA-lncRNA interactions.
• The multi-view fusion and parallel processing approaches overcome limitations of existing methods.
• This model can advance research in gene regulation and disease mechanisms.