Predict potential miRNA-disease associations based on bounded nuclear norm regularization

Predict potential miRNA-disease associations based on bounded nuclear norm regularization

 

 

2023/12/8 16:00:57

Predicting potential miRNA-disease associations is a challenging task in bioinformatics and computational biology. One approach that has been proposed for this task is the use of bounded nuclear norm regularization.

Bounded nuclear norm regularization is a method used to estimate the association between miRNAs and diseases by integrating multiple types of data, such as miRNA expression profiles, disease similarity measures, and known miRNA-disease associations. The key idea behind this method is to model the association matrix as a low-rank matrix, where the rows represent miRNAs, the columns represent diseases, and the entries represent the strength of the association.

The bounded nuclear norm regularization aims to factorize the association matrix into two low-rank matrices, one representing miRNA similarities and the other representing disease similarities. By incorporating the bounded nuclear norm regularization term, which enforces the low-rank property, the method can effectively capture the underlying associations between miRNAs and diseases.

To predict potential miRNA-disease associations using this method, the association matrix is factorized, and the missing entries are estimated based on the low-rank approximation. The predicted values can then be used to identify potential miRNA-disease associations.

It's worth noting that while bounded nuclear norm regularization has shown promising results in predicting miRNA-disease associations, it's just one of many approaches in this field. There are several other methods and algorithms available that utilize different strategies and data sources for prediction. Researchers continue to explore and develop new techniques to improve the accuracy and reliability of miRNA-disease association predictions.