Comparative analysis of RNA-seq data from polyA RNAs selection and ribosomal RNAs deletion protocol by strand-specific RNA sequencing technology

Abstract

The conventional non-strand-specific RNA-seq method is widely used for many studies, but it cannot characterize which strand was the transcript originally came from. Strand-specific RNA library construction methods have been developed to overcome this drawback. Here, we compared transcriptomics data from two mainstream RNA enrichment methods (polyA RNAs selection and ribosomal RNAs deletion) by strand-specific RNA sequencing. Using paired-end strategy, we obtained 175 and 149 million high quality reads without ribosomal RNA reads by ribosomal RNAs deletion and poly(A)+ RNAs selection protocol, respectively. From these reads, rmRNA-seq had lower (53.28%) unique mapping rate than the mRNA-seq (73.89%). But, the ribosomal RNAs deletion protocol detected more known non-coding RNAs, particularly lncRNAs, pseudogenes and snoRNAs. Larger proportion (66.7%) of reads mapping to intronic and intergenic regions in ribosomal RNAs deletion method and fewer percentages (33.3%) of reads aligning to exonic regions compared with poly(A)+ RNAs selection method (35.8% and 64.2%). The ribosomal RNAs deletion protocol provides advantages over the poly(A)+ RNAs selection method in sense-antisense pairs detection. In conclusion, the comparison of these two rRNA enrichment methods provides us insight for utility of each protocol. Moreover, we believe that ribosomal RNAs deletion based strand-specific RNA sequencing show us a more comprehensive view of eukaryotic transcriptomes.