circGPAcorr: an integrative tool for functional annotation of circular RNAs using expression data.

Area of Science:

• Bioinformatics
• Genomics
• Molecular Biology

Background:

• Circular RNAs (circRNAs) are vital in cell development and disease biomarker discovery.
• The functions of many circRNAs remain uncharacterized.
• Inferring circRNA function relies on interactions with microRNAs (miRNAs) and messenger RNAs (mRNAs).

Purpose of the Study:

• To address limitations of existing tools like circGPA, which suffer from sparse validation data and high false positive rates.
• To develop an enhanced algorithm, circGPAcorr, for more precise circRNA functional annotation.
• To improve the accuracy of predicting circRNA-disease associations.

Main Methods:

• Proposed circGPAcorr, an extension of circGPA, incorporating expression data to weight RNA interactions.
• Utilized a generating-polynomial-based method for p-value calculation, assessing result significance.
• Demonstrated the computational difficulty of the problem, identifying it as #P-hard.

Main Results:

• circGPAcorr significantly improves the precision of circRNA functional annotation by weighting interactions with expression data.
• The algorithm was successfully tested on myelodysplastic syndromes expression data, yielding gene ontology annotations consistent with existing literature.
• Validated the algorithm's performance in circRNA-disease association prediction.

Conclusions:

• circGPAcorr offers a more accurate approach to functional annotation of circRNAs compared to previous methods.
• The algorithm provides valuable insights into circRNA roles in diseases, particularly myelodysplastic syndromes.
• Expression data integration is crucial for enhancing the reliability of circRNA interaction predictions and functional annotations.