Union Exon Based Approach for RNA-Seq Gene Quantification: To Be or Not to Be?

Area of Science:

• Bioinformatics
• Genomics
• Molecular Biology

Background:

• RNA-sequencing (RNA-seq) is a key technology for gene expression analysis.
• Reads Per Kilobase per Million reads (RPKM) is commonly used for mRNA abundance estimation.
• Gene quantification methods are broadly categorized into transcript-based and 'union exon'-based approaches.

Purpose of the Study:

• To compare the accuracy of 'union exon'-based and transcript-based methods for RNA-seq gene quantification.
• To evaluate the impact of different quantification approaches on gene expression levels and differential analysis.
• To determine if the 'union exon'-based approach is a suitable practice for RNA-seq gene quantification.

Main Methods:

• Side-by-side comparison of 'union exon'-based and transcript-based RNA-seq gene quantification.
• Analysis of gene and transcript level differential expression.
• Evaluation of factors influencing the discrepancy between the two quantification methods.

Main Results:

• The 'union exon'-based approach significantly underestimates gene expression levels compared to the transcript-based method.
• Average RPKM from 'union exons' was less than 50% of the mean expression from transcript-based analysis.
• The discrepancy is influenced by the number of transcripts per gene, with isoform-level analysis revealing more insights like isoform switches.

Conclusions:

• The 'union exon'-based approach is not recommended for gene quantification in RNA-seq due to significant underestimation of expression.
• Transcript-based quantification provides more biologically meaningful and accurate results.
• Incorporating read coverage and exon-exon spanning reads into algorithms like EM can improve isoform quantification accuracy.