Table_3_Functional Annotation of the Transcriptome of the Pig, Sus scrofa, Based Upon Network Analysis of an RNAseq Transcriptional Atlas.xlsx

The domestic pig (Sus scrofa) is both an economically important livestock species and a model for biomedical research. Two highly contiguous pig reference genomes have recently been released. To support functional annotation of the pig genomes and comparative analysis with large human transcriptomic data sets, we aimed to create a pig gene expression atlas. To achieve this objective, we extended a previous approach developed for the chicken. We downloaded RNAseq data sets from public repositories, down-sampled to a common depth, and quantified expression against a reference transcriptome using the mRNA quantitation tool, Kallisto. We then used the network analysis tool Graphia to identify clusters of transcripts that were coexpressed across the merged data set. Consistent with the principle of guilt-by-association, we identified coexpression clusters that were highly tissue or cell-type restricted and contained transcription factors that have previously been implicated in lineage determination. Other clusters were enriched for transcripts associated with biological processes, such as the cell cycle and oxidative phosphorylation. The same approach was used to identify coexpression clusters within RNAseq data from multiple individual liver and brain samples, highlighting cell type, process, and region-specific gene expression. Evidence of conserved expression can add confidence to assignment of orthology between pig and human genes. Many transcripts currently identified as novel genes with ENSSSCG or LOC IDs were found to be coexpressed with annotated neighbouring transcripts in the same orientation, indicating they may be products of the same transcriptional unit. The meta-analytic approach to utilising public RNAseq data is extendable to include new data sets and new species and provides a framework to support the Functional Annotation of Animals Genomes (FAANG) initiative.

家猪（Sus scrofa）既是兼具重要经济价值的畜禽物种，也是生物医学研究的模式生物。近期已有两份连续性极佳的猪参考基因组被发布。为支撑猪基因组的功能注释，以及开展与大规模人类转录组数据集的比较分析，本研究旨在构建猪基因表达图谱。为达成该目标，我们拓展了此前针对鸡开发的分析方法。我们从公共数据库下载RNA测序（RNA-seq）数据集，将其下采样至统一测序深度，并借助mRNA定量工具Kallisto对参考转录组进行表达量定量。随后我们使用网络分析工具Graphia，在合并后的数据集内鉴定出共表达的转录本簇。基于共表达关联推定原则（guilt-by-association），我们鉴定出了具有高度组织或细胞类型特异性的共表达簇，其中包含此前被证实与细胞谱系决定相关的转录因子。其余簇则富集了与细胞周期、氧化磷酸化等生物学过程相关的转录本。我们采用相同方法，对多份独立肝脏和大脑样本的RNA测序数据进行分析，鉴定出共表达簇，从而揭示了细胞类型、生物学过程以及区域特异性的基因表达特征。表达保守性的证据可提升猪与人类基因间直系同源基因注释的可信度。当前以ENSSSCG或LOC编号标注的诸多新转录本，被发现与同方向相邻的已注释转录本存在共表达现象，这表明它们可能属于同一转录单元的产物。本研究利用公共RNA测序数据的元分析方法可拓展至新数据集与新物种，同时为支撑动物基因组功能注释（Functional Annotation of Animals Genomes, FAANG）计划提供了一套研究框架。