Sequencing the entire RNA molecule leads to a better understanding of the transcriptome architecture. SMARTer (Switching Mechanism at 5′-End of RNA Template) is a technology aimed at generating full-length cDNA from low amounts of mRNA for sequencing by short-read sequencers such as those from Illumina. However, short read sequencing such as Illumina technology includes fragmentation that results in bias and information loss. Here, we built a pipeline, UNAGI or UNAnnotated Gene Identifier, to process long reads obtained with nanopore sequencing and compared this pipeline with the standard Illumina pipeline by studying the Saccharomyces cerevisiae transcriptome in full-length cDNA samples generated from two different biological samples: haploid and diploid cells. Additionally, we processed the long reads with another long read tool, FLAIR. Our strand-aware method revealed significant differential gene expression that was masked in Illumina data by antisense transcripts. Our pipeline, UNAGI, outperformed the Illumina pipeline and FLAIR in transcript reconstruction (sensitivity and specificity of 80% and 40% vs. 18% and 34% and 79% and 32%, respectively). Moreover, UNAGI discovered 3877 unannotated transcripts including 1282 intergenic transcripts while the Illumina pipeline discovered only 238 unannotated transcripts. For isoforms profiling, UNAGI also outperformed the Illumina pipeline and FLAIR in terms of sensitivity (91% vs. 82% and 63%, respectively). But the low accuracy of nanopore sequencing led to a closer gap in terms of specificity with Illumina pipeline (70% vs. 63%) and to a huge gap with FLAIR (70% vs 0.02%).

中文翻译：

---

UNAGI：用于纳米孔全长 cDNA 测序的自动化管道揭示了酵母中的新转录本和亚型。

对整个 RNA 分子进行测序可以更好地理解转录组结构。SMARTer（RNA 模板 5' 端的切换机制）是一项旨在从少量 mRNA 生成全长 cDNA 的技术，以便通过短读长测序仪（例如 Illumina 的测序仪）进行测序。然而，诸如 Illumina 技术之类的短读长测序包含碎片，会导致偏差和信息丢失。在这里，我们构建了一个管道，UNAGI 或 UNAnnotated Gene Identifier，用于处理通过纳米孔测序获得的长读长，并通过研究从两种不同生物样本生成的全长 cDNA 样本中的酿酒酵母转录组，将该管道与标准 Illumina 管道进行比较：和二倍体细胞。此外，我们使用另一个长读工具 FLAIR 处理长读。我们的链感知方法揭示了在 Illumina 数据中被反义转录本掩盖的显着差异基因表达。我们的流程 UNAGI 在转录本重建方面优于 Illumina 流程和 FLAIR（灵敏度和特异性分别为 80% 和 40% vs. 18% 和 34% 以及 79% 和 32%）。此外，UNAGI 发现了 3877 个未注释的转录本，其中包括 1282 个基因间转录本，而 Illumina 管道仅发现了 238 个未注释的转录本。对于异构体分析，UNAGI 在灵敏度方面也优于 Illumina 管道和 FLAIR（分别为 91% vs. 82% 和 63%）。但纳米孔测序的准确度较低，导致其与 Illumina 流程的特异性差距更小（70% vs. 63%），与 FLAIR 的差距巨大（70% vs 0.02%）。