LDCMFC: Predicting Long Non-Coding RNA and Disease Association Using Collaborative Matrix Factorization Based on...

Area of Science:

• Bioinformatics and computational biology
• Genomics and molecular biology
• Disease association studies

Background:

• Long non-coding RNAs (lncRNAs) play a crucial role in various intractable diseases.
• Identifying lncRNA-disease associations (LDAs) is challenging and expensive, limiting current knowledge.
• Accurate and efficient methods are vital for discovering novel LDAs.

Purpose of the Study:

• To propose a novel computational method for identifying potential lncRNA-disease associations (LDAs).
• To enhance the robustness and accuracy of LDA prediction algorithms.
• To provide a feasible tool for exploring unknown LDAs.

Main Methods:

• Developed a collaborative matrix factorization method based on correntropy (LDCMFC).
• Replaced Euclidean distance minimization with correntropy maximization for improved robustness.
• Utilized the weighted K nearest known neighbor (WKNKN) method to reconstruct the adjacency matrix.

Main Results:

• LDCMFC achieved a high Area Under the Curve (AUC) of 0.8628 in 5-fold cross-validation.
• The method demonstrated superior performance compared to traditional LDA prediction techniques.
• Experimental validation in three cancer types confirmed the accuracy of predicted lncRNAs.

Conclusions:

• LDCMFC is a feasible and effective computational method for predicting lncRNA-disease associations.
• The proposed approach offers a robust and accurate alternative for LDA discovery.
• This method can aid in understanding the role of lncRNAs in intractable diseases.