AtRTD2: A Reference Transcript Dataset for accurate quantification of alternative splicing and expression changes in Arabidopsis thaliana RNA-seq data
AtRTD2: A Reference Transcript Dataset for accurate quantification of alternative splicing and expression changes in Arabidopsis thaliana RNA-seq data
复制标题
DOI：
10.1101/051938
复制
发表时间：
2016-05
期刊：
影响因子：
--
通讯作者：
Runxuan Zhang;C. Calixto;Yamile Marquez;Peter Venhuizen;Nikoleta A Tzioutziou;Wenbin Guo;Mark A. Spensley;Nicolas Frei dit Frey;H. Hirt;A. James;H. G. Nimmo;A. Barta;M. Kalyna;John W. S. Brown
中科院分区：
文献类型：
--
作者：
Runxuan Zhang;C. Calixto;Yamile Marquez;Peter Venhuizen;Nikoleta A Tzioutziou;Wenbin Guo;Mark A. Spensley;Nicolas Frei dit Frey;H. Hirt;A. James;H. G. Nimmo;A. Barta;M. Kalyna;John W. S. Brown
背景选择性剪接是调节基因表达的主要转录后机制，并影响大多数真核生物中的广泛过程和反应。 RNA 测序 (RNA-seq) 可以对单个转录亚型进行全基因组定量，以识别表达和选择性剪接的变化。 RNA-seq 是一种重要的现代工具，但其准确量化转录亚型的能力取决于转录信息的多样性、完整性和质量。结果我们开发了一个新的拟南芥参考转录本数据集 (AtRTD2)，用于 RNA-seq 分析，其中包含超过 82k 个非冗余转录本，其中 74,194 个转录本来自 27,667 个蛋白质编码基因。 AtRTD2 中总共 13,524 个蛋白质编码基因具有至少一个选择性剪接转录本，因此拟南芥中 22,453 个含有内含子的蛋白质编码基因中约 60% 经历选择性剪接。在 2,000 多个转录本中鉴定出了 600 多个假定的 U12 内含子。 AtRTD2 是从 ca 的转录本组装生成的。来自 129 个 RNA-seq 文库的 285 个 RNA-seq 数据集的 85 亿对读数，并与之前的版本、AtRTD 和 Araport11 转录本组件合并。 AtRTD2 增加了转录本的多样性，并通过应用严格的过滤器代表了迄今为止拟南芥属最广泛和最准确的转录本集合。我们已经证明了 Salmon 分析的 RNA-seq 数据和高分辨率 RT-PCR 实验数据的选择性剪接比率总体上具有良好的相关性。然而，我们观察到具有多个转录本的基因的转录同工型定量不准确，这些转录本的 UTR 长度存在变化。这种变异在 RNA-seq 分析程序中并未得到有效纠正，因此通常会影响 RNA-seq 分析。为了解决这个问题，我们测试了 AtRTD2 的不同全基因组修饰，以改进转录本定量和选择性剪接分析。因此，我们发布了专门用于选择性剪接异构体定量的 AtRTD2-QUASI，并证明它优于其他可用的 RNA-seq 分析转录组。结论 我们已经生成了用于拟南芥 RNA 序列分析的新转录组资源 (AtRTD2)，旨在解决基因表达研究中不同亚型和选择性剪接的量化问题。选择性剪接变化的实验验证发现，由于 UTR 长度变化导致转录本定量不准确。为了解决这个问题，我们还发布了修改后的参考转录组AtRTD2-QUASI，用于转录亚型的定量，其与实验数据显示出高度相关性。
Background Alternative splicing is the major post-transcriptional mechanism by which gene expression is regulated and affects a wide range of processes and responses in most eukaryotic organisms. RNA-sequencing (RNA-seq) can generate genome-wide quantification of individual transcript isoforms to identify changes in expression and alternative splicing. RNA-seq is an essential modern tool but its ability to accurately quantify transcript isoforms depends on the diversity, completeness and quality of the transcript information. Results We have developed a new Reference Transcript Dataset for Arabidopsis (AtRTD2) for RNA-seq analysis containing over 82k non-redundant transcripts, whereby 74,194 transcripts originate from 27,667 protein-coding genes. A total of 13,524 protein-coding genes have at least one alternatively spliced transcript in AtRTD2 such that about 60% of the 22,453 protein-coding, intron-containing genes in Arabidopsis undergo alternative splicing. More than 600 putative U12 introns were identified in more than 2,000 transcripts. AtRTD2 was generated from transcript assemblies of ca. 8.5 billion pairs of reads from 285 RNA-seq data sets obtained from 129 RNA-seq libraries and merged along with the previous version, AtRTD, and Araport11 transcript assemblies. AtRTD2 increases the diversity of transcripts and through application of stringent filters represents the most extensive and accurate transcript collection for Arabidopsis to date. We have demonstrated a generally good correlation of alternative splicing ratios from RNA-seq data analysed by Salmon and experimental data from high resolution RT-PCR. However, we have observed inaccurate quantification of transcript isoforms for genes with multiple transcripts which have variation in the lengths of their UTRs. This variation is not effectively corrected in RNA-seq analysis programmes and will therefore impact RNA-seq analyses generally. To address this, we have tested different genome-wide modifications of AtRTD2 to improve transcript quantification and alternative splicing analysis. As a result, we release AtRTD2-QUASI specifically for use in Quantification of Alternatively Spliced Isoforms and demonstrate that it out-performs other available transcriptomes for RNA-seq analysis. Conclusions We have generated a new transcriptome resource for RNA-seq analyses in Arabidopsis (AtRTD2) designed to address quantification of different isoforms and alternative splicing in gene expression studies. Experimental validation of alternative splicing changes identified inaccuracies in transcript quantification due to UTR length variation. To solve this problem, we also release a modified reference transcriptome, AtRTD2-QUASI for quantification of transcript isoforms, which shows high correlation with experimental data.