Division of Biological Science, Faculty of Science, Prince of Songkla University, Hat Yai, Songkhla, Thailand.
Center for Genomics and Bioinformatics Research, Faculty of Science, Prince of Songkla University, Hat Yai, Songkhla, Thailand.
PeerJ. 2022 Nov 10;10:e14344. doi: 10.7717/peerj.14344. eCollection 2022.
Differential gene expression analysis using RNA sequencing technology (RNA-Seq) has become the most popular technique in transcriptome research. Although many R packages have been developed to analyze differentially expressed genes (DEGs), several evaluations have shown that no single DEG analysis method outperforms all others. The validity of DEG identification could be increased by using multiple methods and producing the consensus results. However, DEG analysis methods are complex and most of them require prior knowledge of a programming language or command-line shell. Users who do not have this knowledge need to invest time and effort to acquire it.
We developed a novel web application called "bestDEG" to automatically analyze DEGs with different tools and compare the results. A differential expression (DE) analysis pipeline was created combining the edgeR, DESeq2, NOISeq, and EBSeq packages; selected because they use different statistical methods to identify DEGs. bestDEG was evaluated on human datasets from the MicroArray Quality Control (MAQC) project.
The performance of the bestDEG web application with the human datasets showed excellent results, and the consensus method outperformed the other DE analysis methods in terms of precision (94.71%) and specificity (97.01%). bestDEG is a rapid and efficient tool to analyze DEGs. With bestDEG, users can select DE analysis methods and parameters in the user-friendly web interface. bestDEG also provides a Venn diagram and a table of results. Moreover, the consensus method of this tool can maximize the precision or minimize the false discovery rate (FDR), which reduces the cost of gene expression validation by minimizing wet-lab experiments.
使用 RNA 测序技术（RNA-Seq）进行差异基因表达分析已成为转录组研究中最流行的技术。尽管已经开发了许多 R 包来分析差异表达基因（DEG），但多项评估表明，没有一种单一的 DEG 分析方法优于所有其他方法。通过使用多种方法并产生共识结果，可以提高 DEG 鉴定的有效性。然而，DEG 分析方法复杂，大多数方法需要编程语言或命令行外壳的先验知识。没有这些知识的用户需要投入时间和精力来学习。
我们开发了一个名为“bestDEG”的新型 Web 应用程序，用于使用不同的工具自动分析 DEG 并比较结果。创建了一个差异表达（DE）分析管道，结合了 edgeR、DESeq2、NOISeq 和 EBSeq 包；选择它们是因为它们使用不同的统计方法来识别 DEG。在 MicroArray Quality Control（MAQC）项目的人类数据集上评估了 bestDEG。
bestDEG 网络应用程序在人类数据集上的性能表现出色，共识方法在精度（94.71%）和特异性（97.01%）方面均优于其他 DE 分析方法。bestDEG 是一种快速有效的分析 DEG 的工具。使用 bestDEG，用户可以在用户友好的 Web 界面中选择 DE 分析方法和参数。bestDEG 还提供了一个 Venn 图和一个结果表。此外，该工具的共识方法可以最大化精度或最小化错误发现率（FDR），从而通过最小化湿实验室实验降低基因表达验证的成本。
