Reconstruction of full-length circular RNAs enables isoform-level quantification

Area of Science:

• Molecular Biology
• Bioinformatics
• Genomics

Background:

• Current circRNA research focuses on biological functions, but is hindered by challenges in full-length sequence determination and accurate isoform-level quantification.
• Existing methods for circRNA detection primarily rely on back-splice junction (BSJ) analysis, which may miss low-abundance circRNAs.

Purpose of the Study:

• To develop a novel approach for circRNA detection, full-length sequence reconstruction, and isoform-level quantification.
• To improve the identification of low-abundance circRNAs.
• To highlight the importance of isoform-level analysis in circRNA differential expression studies.

Main Methods:

• Proposed a new feature, reverse overlap (RO), for circRNA detection.
• Combined RO and BSJ features to develop a novel computational approach.
• Utilized the CIRI-full software for analysis.
• Compared BSJ-level and isoform-level differential expression analyses in human liver tumor and normal tissues.

Main Results:

• The reverse overlap (RO) feature demonstrated superior performance in identifying low-abundance circRNAs compared to BSJ-based methods.
• The combined RO and BSJ approach enabled effective reconstruction of full-length circRNAs.
• Isoform-level quantification provided deeper insights into circRNA expression patterns.
• Differential expression analysis revealed significant differences between BSJ-level and isoform-level results, underscoring the necessity of isoform resolution.

Conclusions:

• The novel approach combining RO and BSJ features significantly enhances circRNA detection, reconstruction, and quantification.
• Isoform-level resolution is critical for comprehensive understanding of circRNA biological functions and differential expression.
• The CIRI-full software provides a valuable tool for advancing circRNA research to isoform-level resolution.