Replicate sequencing libraries are important for quantification of allelic imbalance

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Allelic imbalance (AI) analysis in RNA sequencing (RNA-seq) is crucial for studying transcriptional regulation in diploid organisms.
• Current practices often rely on single RNA-seq libraries, potentially leading to inaccurate AI quantification due to unaddressed technical noise.

Purpose of the Study:

• To demonstrate the limitations of single-library RNA-seq for reliable AI analysis.
• To introduce Qllelic, a computational method that accurately accounts for technical noise using replicate RNA-seq libraries.
• To improve the accuracy and reproducibility of allele-specific expression analysis.

Main Methods:

• Theoretical and experimental evaluation of AI signal from single versus replicate RNA-seq libraries.
• Development and application of the Qllelic computational approach utilizing replicate libraries.
• Analysis of existing and new datasets to assess Qllelic's performance.

Main Results:

• Single RNA-seq libraries provide insufficient data for reliable quantification of technical noise in AI.
• Qllelic significantly reduces the false positive rate in allele-specific analysis by accurately modeling technical noise.
• The Qllelic approach enhances the reproducibility of AI estimates.

Conclusions:

• Replicate RNA-seq libraries are essential for robust AI quantification and accurate allele-specific expression analysis.
• Qllelic offers a reliable computational solution for accounting for technical variation in AI studies.
• The study provides insights into designing RNA-seq experiments for transcriptome-wide AI quantification and differential analysis.