AppleMDO: A Multi-Dimensional Omics Database for Apple Co-Expression Networks and Chromatin States
AppleMDO: A Multi-Dimensional Omics Database for Apple Co-Expression Networks and Chromatin States
复制标题
AppleMDO：苹果共表达网络和染色质状态的多维组学数据库
DOI：
10.3389/fpls.2019.01333
复制
发表时间：
2019-10-22
影响因子：
5.6
通讯作者：
Su, Zhen
中科院分区：
文献类型：
--
作者：
Da, Lingling;Liu, Yue;Su, Zhen
苹果是世界上温带地区栽培最多的果树之一，是一种重要的经济作物。近年来，大量高质量的苹果转录组和表观基因组数据库被公开，这将有助于推断基因调控关系，从而在基因组水平上预测基因功能。通过整合苹果基因组、转录组和表观基因组数据集，我们构建了共表达网络，鉴定了功能模块，并预测了染色质状态。共整合了112个RNA-seq数据集，构建了一个全局网络和一个条件网络（组织优先网络）。此外，共鉴定了1，076个具有密切相关基因集的功能模块，以评估生物网络的模块性，并进一步进行功能富集分析。结果表明，许多模块的功能与发育、次生代谢、激素反应和转录调控有关。转录调控与染色质上的表观遗传标记密切相关。共20个表观基因组数据集，其中包括ChIP-seq，DNase-seq，和DNA甲基化分析数据集，被整合并用于分类染色质状态。基于ChromHMM算法，将基因组划分为620,122个片段，根据表观遗传标记和富集特征区域的组合将片段分为24种状态。最后，通过对不同组学数据集的协同分析，建立了AppleMDO（http：//bioinformatics.cau.edu.cn/AppleMDO/）在线数据库，用于交叉引用和探索苹果基因可能的新功能。此外，还提供了基因注释信息和功能支持工具包。我们的数据库可能方便研究人员开发的重要农艺性状相关基因的功能的见解，并可能作为其他果树的参考。
As an economically important crop, apple is one of the most cultivated fruit trees in temperate regions worldwide. Recently, a large number of high-quality transcriptomic and epigenomic datasets for apple were made available to the public, which could be helpful in inferring gene regulatory relationships and thus predicting gene function at the genome level. Through integration of the available apple genomic, transcriptomic, and epigenomic datasets, we constructed co-expression networks, identified functional modules, and predicted chromatin states. A total of 112 RNA-seq datasets were integrated to construct a global network and a conditional network (tissue-preferential network). Furthermore, a total of 1,076 functional modules with closely related gene sets were identified to assess the modularity of biological networks and further subjected to functional enrichment analysis. The results showed that the function of many modules was related to development, secondary metabolism, hormone response, and transcriptional regulation. Transcriptional regulation is closely related to epigenetic marks on chromatin. A total of 20 epigenomic datasets, which included ChIP-seq, DNase-seq, and DNA methylation analysis datasets, were integrated and used to classify chromatin states. Based on the ChromHMM algorithm, the genome was divided into 620,122 fragments, which were classified into 24 states according to the combination of epigenetic marks and enriched-feature regions. Finally, through the collaborative analysis of different omics datasets, the online database AppleMDO (http://bioinformatics.cau.edu.cn/AppleMDO/) was established for cross-referencing and the exploration of possible novel functions of apple genes. In addition, gene annotation information and functional support toolkits were also provided. Our database might be convenient for researchers to develop insights into the function of genes related to important agronomic traits and might serve as a reference for other fruit trees.