Quantifying circular RNA expression from RNA-seq data using model-based framework.

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Circular RNAs (circRNAs) are non-coding RNAs with cell type- and tissue-specific expression patterns.
• Their precise functions are under investigation, but expression patterns suggest critical roles in biological processes.
• Accurate quantification of circRNA expression from RNA-seq data is essential but challenging with existing methods.

Purpose of the Study:

• To develop a novel computational strategy for accurate circRNA expression quantification from RNA-seq data.
• To improve the accuracy of both circular and linear RNA expression estimation.
• To provide a freely available software tool for circRNA analysis.

Main Methods:

• Transformed circular transcripts into pseudo-linear transcripts.
• Utilized the Sailfish algorithm to estimate expression values for both circular and linear transcripts.
• Developed a program named Sailfish-cir to implement the strategy.

Main Results:

• The new strategy accurately estimates expression for both linear and circular transcripts.
• Quantification performance is influenced by gene length, expression level, and circRNA/linear RNA ratio.
• Sailfish-cir demonstrated superior performance compared to count-based tools on simulated and real RNA-seq datasets.
• Incorporating circRNAs enhanced the accuracy of linear transcript quantification.

Conclusions:

• Sailfish-cir offers a robust and accurate method for circRNA expression quantification.
• The tool improves the accuracy of both circRNA and linear RNA quantification from RNA-seq data.
• This advancement aids in understanding the biological roles of circRNAs.