Predicting Human lncRNA-Disease Associations Based on Geometric Matrix Completion.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Long non-coding RNAs (lncRNAs) are increasingly implicated in various diseases.
• Existing experimentally verified lncRNA-disease associations are limited, hindering diagnosis, treatment, and mechanistic studies.
• Current computational methods face challenges in precise prediction and leveraging intrinsic data structures.

Purpose of the Study:

• To develop a novel computational method, GMCLDA (Geometric Matrix Completion lncRNA-Disease Association), for inferring lncRNA-disease associations.
• To improve the accuracy and robustness of lncRNA-disease association prediction.
• To facilitate the identification of potential lncRNA biomarkers for diseases.

Main Methods:

• GMCLDA utilizes geometric matrix completion, integrating functional similarity of lncRNAs and phenotypic similarity of diseases.
• Disease semantic similarity is computed using the Disease Ontology (DO) hierarchy.
• lncRNA sequence similarity is measured via the Needleman-Wunsch algorithm, and Gaussian interaction profile kernel similarity is employed.

Main Results:

• GMCLDA demonstrates superior accuracy in predicting lncRNA-disease associations compared to state-of-the-art methods.
• The method effectively utilizes the intrinsic structure of the association matrix through geometric matrix completion.
• Case studies successfully identified potential lncRNAs associated with renal cancer.

Conclusions:

• GMCLDA offers a robust and accurate approach for predicting lncRNA-disease associations.
• The method enhances the prioritization of potential associations for diagnostic and therapeutic applications.
• GMCLDA contributes to a deeper understanding of disease mechanisms at the lncRNA level.