Transcriptome analysis by strand-specific sequencing of complementary DNA

High throughput complementary DNA sequencing (RNA-Seq) is a powerful tool for whole-transcriptome analysis, supplying information about a transcript’s expression level and structure. However, RNA-Seq is unable to determine the polarity of transcripts, and therefore cannot identify which strand has been transcribed. Also, it cannot distinguish whether one or both strands are being transcribed, for example with antisense transcripts reported to be associated with more than half of the genes in eukaryotes, or with overlapping genes that are abundant in compact genomes of prokaryotes and lower eukaryotes. Here, we present a cDNA sequencing protocol that preserves information about a transcript’s direction. Using S. cerevisiae and mouse brain transcriptomes as models, we demonstrate that knowing the transcript’s orientation allows more accurate determination of the structure and expression of genes. It also helps to identify new genes and enables studying promoter-associated and antisense transcription. The transcriptional landscapes we obtained are available online at http://genseq.molgen.mpg.de/ssRNA/.