PSnoD: identifying potential snoRNA-disease associations based on bounded nuclear norm regularization.

Area of Science:

• Genomics
• Computational Biology
• Disease Association Studies

Background:

• Small nucleolar RNAs (snoRNAs) are implicated in human complex diseases.
• Identifying snoRNA-disease associations aids disease pathogenesis understanding.
• Traditional experimental methods for association discovery are resource-intensive.

Purpose of the Study:

• To propose a novel computational method, PSnoD, for predicting snoRNA-disease associations.
• To evaluate the performance and efficiency of PSnoD against existing methods.

Main Methods:

• Developed PSnoD, a bounded nuclear norm regularization-based prediction method.
• Utilized 5-fold stratified shuffle split for benchmark experiments.
• Compared computational efficiency with other matrix completion techniques.

Main Results:

• PSnoD demonstrated superior performance compared to state-of-the-art methods.
• Achieved an Area Under the Receiver Operating Characteristic curve (AUROC) of 0.90 and Area Under the Precision-Recall curve (AUPR) of 0.55.
• Case studies confirmed PSnoD's capability in screening potential snoRNA-disease associations.

Conclusions:

• PSnoD is an effective and computationally efficient tool for predicting snoRNA-disease associations.
• The method aids in identifying potential links, advancing disease research.
• A publicly accessible web server and code are available for PSnoD.