Song Bowen, Tang Yujiao, Wei Zhen, Liu Gang, Su Jionglong, Meng Jia, Chen Kunqi
Department of Biological Sciences, Xi'an Jiaotong-Liverpool University, Suzhou, China.
Institute of Integrative Biology, University of Liverpool, Liverpool, United Kingdom.
Front Genet. 2020 Mar 12;11:88. doi: 10.3389/fgene.2020.00088. eCollection 2020.
Known as the "fifth RNA nucleotide", pseudouridine (Ψ or psi) is the first-discovered and most abundant RNA modification occurring at the Uridine site, and it plays a prominent role in a number of biological processes. Thousands of Ψ sites have been identified within different biological contexts thanks to the advancement in high-throughput sequencing technology; nevertheless, the transcriptome-wide distribution, biomolecular functions, regulatory mechanisms, and disease relevance of pseudouridylation are largely elusive. We report here a web server-PIANO-for seudouridine site (Ψ) dentification nd fuctional anntation. PIANO was built upon a high-accuracy predictor that takes advantage of both conventional sequence features and 42 additional genomic features. When tested on six independent datasets generated from four independent Ψ-profiling technologies (Ψ-seq, RBS-seq, Pseudo-seq, and CeU-seq) as benchmarks, PIANO achieved an average AUC of 0.955 and 0.838 under the full transcript and mature mRNA models, respectively, marking a substantial improvement in accuracy compared to the existing Ψ-site prediction methods, i.e., PPUS (0.713 and 0.707), iRNA-PseU (0.713 and 0.712), and PseUI (0.634 and 0.652). Besides, PIANO web server systematically annotates the predicted Ψ sites with post-transcriptional regulatory mechanisms (miRNA-targets, RBP-binding regions, and splicing sites) in its prediction report to help the users explore potential machinery of Ψ. Moreover, a concise query interface was also built for 4,303 known Ψ sites, which is currently the largest collection of experimentally validated human Ψ sites. The PIANO website is freely accessible at: http://piano.rnamd.com.
假尿苷（Ψ 或 psi）被称为“第五种 RNA 核苷酸”，是最早发现且在尿苷位点出现最为丰富的 RNA 修饰，它在许多生物学过程中发挥着重要作用。由于高通量测序技术的进步，在不同生物学背景下已鉴定出数千个 Ψ 位点；然而，假尿苷化在全转录组范围内的分布、生物分子功能、调控机制以及与疾病的相关性在很大程度上仍不清楚。我们在此报告一个用于假尿苷位点（Ψ）鉴定和功能注释的网络服务器——PIANO。PIANO 基于一个高精度预测器构建，该预测器利用了传统序列特征和另外 42 个基因组特征。当以从四种独立的 Ψ 谱分析技术（Ψ - seq、RBS - seq、Pseudo - seq 和 CeU - seq）生成的六个独立数据集作为基准进行测试时，PIANO 在完整转录本模型和成熟 mRNA 模型下分别实现了平均 AUC 为 0.955 和 0.838，与现有的 Ψ 位点预测方法（即 PPUS（0.713 和 0.707）、iRNA - PseU（0.713 和 0.712）以及 PseUI（0.634 和 0.652）相比，准确性有了显著提高。此外，PIANO 网络服务器在其预测报告中用转录后调控机制（miRNA 靶标、RBP 结合区域和剪接位点）对预测的 Ψ 位点进行系统注释，以帮助用户探索 Ψ 的潜在机制。此外，还为 4303 个已知的 Ψ 位点构建了一个简洁的查询界面，这是目前最大的经过实验验证的人类 Ψ 位点集合。PIANO 网站可免费访问：http://piano.rnamd.com。
