Traditional versus 3′ RNA-seq in a non-model species

One limitation of the widely used RNA-seq method is that long transcripts are represented by more reads than shorter transcripts, resulting in a biased estimation of expression levels. The 3′ RNA-seq method, which yields only one sequence per transcript, bypasses this limitation. Here, RNA was extracted from two samples, in which we expected to find differentially expressed genes. Each was processed by both traditional and 3′ RNA-seq protocols. Both methods yielded similar differentially expressed genes and estimated expression levels in a comparable way, confirming they both represent valid tools for RNA-seq analysis. Notably, however, we identified more differentially expressed transcripts with the 3′ RNA-seq method, suggesting a greater power to detect expression variation using this method. Hence, when little genomic information is available for the species studied, the standard RNA-seq presents a better cost-benefit compromise, whereas for model species, the 3′ RNA-seq method might more accurately detect differential expression.