paperKB
coga / coga-kb
Help
Sign in

Chunk #5 — Results — Quantitation of the Extent of Transcription of the Human Genome

Source
Pervasive transcription of the human genome produces thousands of previously unidentified long intergenic noncoding RNAs.
Embedded
yes

Text

We have analyzed six novel RNA-seq datasets generated as part of the Human Epigenome Atlas (http://www.genboree.org/epigenomeatlas/index.rhtml) and 121 previously published RNA-seq datasets representing 23 human tissues under multiple conditions and consisting of over 4.5 billion uniquely mapped reads (Table S1). This set of RNA-seq data allowed for detection of both rare and tissue-specific transcription events that would otherwise be undetectable. In contrast to the limited reach of prior tiling array studies [2]–[5], we analyzed the much larger portion (83.4%) of the genome to which RNA-seq reads can be uniquely mapped thus providing a broader view of the transcriptome. At a threshold of one RNA-seq read, we observed reads mapping to 78.9% of the genome and, if additional evidence of transcription is taken into account including the full structures of known genes, spliced ESTs and cDNAs, we found evidence that 85.2% of the genome is transcribed (Figure 1A). This result closely agrees with the recently published findings from the ENCODE project in which evidence for transcription of 83.7% of the genome was uncovered [14]. Interestingly, even with 4.5 billion mapped reads,