Normalized long read RNA sequencing in chicken reveals transcriptome complexity similar to human

Abstract

BackgroundDespite the significance of chicken as a model organism, our understanding of the chicken transcriptome is limited compared to human. This issue is common to all non-human vertebrate annotations due to the difficulty in transcript identification from short read RNAseq data. While previous studies have used single molecule long read sequencing for transcript discovery, they did not perform RNA normalization and 5′-cap selection which may have resulted in lower transcriptome coverage and truncated transcript sequences.ResultsWe sequenced normalised chicken brain and embryo RNA libraries with Pacific Bioscience Iso-Seq. 5′ cap selection was performed on the embryo library to provide methodological comparison. From these Iso-Seq sequencing projects, we have identified 60 k transcripts and 29 k genes within the chicken transcriptome. Of these, more than 20 k are novel lncRNA transcripts with ~3 k classified as sense exonic overlapping lncRNA, which is a class that is underrepresented in many vertebrate annotations. The relative proportion of alternative transcription events revealed striking similarities between the chicken and human transcriptomes while also providing explanations for previously observed genomic differences.ConclusionsOur results indicate that the chicken transcriptome is similar in complexity compared to human, and provide insights into other vertebrate biology. Our methodology demonstrates the potential of Iso-Seq sequencing to rapidly expand our knowledge of transcriptomics.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-017-3691-9) contains supplementary material, which is available to authorized users.

New genome assembly of the barn owl (Tyto alba alba)

“…We identified 38,000 protein‐coding genes in the genome of the European barn owl. In the chicken genome version 5, Warren identified 19,119 protein‐coding genes and 6,839 noncoding genes (Warren et al, ) and Kuo 60,000 different transcripts using long‐read RNA sequencing (Kuo et al, ) suggesting that the range of expected genes and transcripts in birds is as high as in mammals. Concerning the comparison of the predicted proteins with BUSCO orthologs, few papers have reported it.…”

“…We identified 38,000 protein-coding genes in the genome of the European barn owl. In the chicken genome version 5, Warren identified 19,119 protein-coding genes and 6,839 noncoding genes (Warren et al, 2017) and Kuo 60,000 different transcripts using long-read RNA sequencing (Kuo et al, 2017) (Tiley, Kimball, Braun, & Burleigh, 2018). This places the European barn owl and its annotation (with up to 74% of protein retrieved), in the upper hand of the annotated avian genomes.…”

Full-length mRNA sequencing uncovers a widespread coupling between transcription and mRNA processing

“…However, as a result of the complexity of human transcriptome and limited size of the sequenced RNA fragment, this approach disappoints in precise reconstruction and reliable expression estimation of transcript variants 6,7,44 . In contrast, single-molecule long-read sequencing provides a unique opportunity to reveal the true complexity of the transcriptome 9,10,45,46 as it can determine the full structure of individual transcripts by full-length sequencing.…”

Avian transcriptomics: opportunities and challenges

“…New technologies such as the Pacific Biosciences (PacBio) Isoform Sequencing (Iso-Seq) application (Gonzalez-Garay 2016), can generate full-length cDNA sequences, and hence reduce the need for reconstructing the full-length RNA molecule by post hoc assembly approaches (Rhoads and Au 2015). The first avian transcriptomes have recently been sequenced using this new technology (Kuo et al 2017;Workman et al 2017). Another new application, which enables sequencing of full-length transcripts, is the MinION from Oxford Nanopore Technologies (Ayub et al 2013;Garalde et al 2016).…”