College of Computer Science and Technology, Jilin University, Changchun, Jilin, China.
Computational Systems Biology Lab, Department of Biochemistry and Molecular Biology, and Institute of Bioinformatics, University of Georgia, GA, USA.
Sci Rep. 2017 Mar 7;7:43925. doi: 10.1038/srep43925.
A transcription unit (TU) consists of K ≥ 1consecutive genes on the same strand of a bacterial genome that are transcribed into a single mRNA molecule under certain conditions. Their identification is an essential step in elucidation of transcriptional regulatory networks. We have recently developed a machine-learning method to accurately identify TUs from RNA-seq data, based on two features of the assembled RNA reads: the continuity and stability of RNA-seq coverage across a genomic region. While good performance was achieved by the method on Escherichia coli and Clostridium thermocellum, substantial work is needed to make the program generally applicable to all bacteria, knowing that the program requires organism specific information. A web server, named SeqTU, was developed to automatically identify TUs with given RNA-seq data of any bacterium using a machine-learning approach. The server consists of a number of utility tools, in addition to TU identification, such as data preparation, data quality check and RNA-read mapping. SeqTU provides a user-friendly interface and automated prediction of TUs from given RNA-seq data. The predicted TUs are displayed intuitively using HTML format along with a graphic visualization of the prediction.
转录单元（TU）由细菌基因组同一条链上的 K≥1 个连续基因组成，在特定条件下可转录成单个 mRNA 分子。在阐明转录调控网络方面，其鉴定是必不可少的一步。我们最近开发了一种基于组装 RNA 读段两个特征的机器学习方法，可从 RNA-seq 数据中准确识别 TU：基因组区域内 RNA-seq 覆盖的连续性和稳定性。该方法在大肠杆菌和热纤维梭菌中表现出良好的性能，但仍需要大量工作使该程序普遍适用于所有细菌，因为该程序需要特定于生物体的信息。名为 SeqTU 的网络服务器使用机器学习方法，针对任何细菌的给定 RNA-seq 数据自动识别 TU。该服务器除 TU 识别外，还包含多个实用工具，例如数据准备、数据质量检查和 RNA 读段映射。SeqTU 提供了一个用户友好的界面，可从给定的 RNA-seq 数据中自动预测 TU。预测的 TU 以 HTML 格式直观显示，并提供预测的图形可视化。
