Proteomic signatures of human visceral and subcutaneous adipocytes - Supplementary files

<b>Supplemental Information</b><b>S File 1 – Dataset with normalized and imputed intensity values</b>MaxQuant search proteinGroups.txt dataset with the following modifications: a) removal of decoy hits and contaminant protein groups; b) exclusion of 4 male sample pairs and an outlying sample pair; b) protein group intensities log₂transformation, c) LoessF normalization, and d) missing values imputation using the imp4p package; e) filtration of protein groups with less than 8 measured intensity values for VA or SA, and protein groups identified to less than 2 peptides within the SA or VA group of samples. The filtered dataset with normalized and imputed intensities was used for the comparative analysis using the Limma R package. <b>S File 2 – Limma differential expression analysis results</b>Differential expression analysis using the LIMMA R package ²⁸. The linear model used to compare paired differences between SA and VA samples was adjusted for batch effect by adding batch number as a variable in the model. The correlation between sample pairs was included in the linear model using appropriate functions from the LIMMA package ²⁹. Subsequently, the results were adjusted for multiple hypothesis testing using the Benjamini and Hochberg procedure ³⁰ implemented in the LIMMA package. <b>S File 3a – SA Reactome over-representation pathway analysis</b>The file was retrieved submitting the list of UniProt accessions of all significantly upregulated SA proteins into the Reactome data analysis tool. <b>S File 3b – VA Reactome over-representation pathway analysis</b>The file was retrieved submitting the list of UniProt accessions of all significantly upregulated VA proteins into the Reactome data analysis tool. <b>S File 4a – Pathway enrichment analysis of all differentially expressed proteins</b>Cytoscape ClueGO plugin Reactome pathways and reactions enrichment analysis results using all upregulated SA and VA proteins submitted as separate groups. The analysis was performed using default settings but showing only results with a p-value &lt; 0.05. <b>S File 4b – Pathway enrichment analysis of SA upregulated proteins</b>Cytoscape ClueGO plugin Reactome pathways and reactions enrichment analysis results for SA upregulated proteins. The analysis was performed using default settings but showing only results with a p-value &lt; 0.05.<b> </b><b>S File 4c – Pathway enrichment analysis of VA upregulated proteins</b>Cytoscape ClueGO plugin Reactome pathways and reactions enrichment analysis results for VA upregulated proteins. The analysis was performed using default settings but showing only results with a p-value &lt; 0.05.<b> </b><b>S File 5 – SignalP prediction of putative secreted proteins</b>The output of putative secreted proteins analysis using SignalP-5.0 Server. This analysis was performed using the FASTA sequence of the most differentially expressed proteins with log₂FC &gt; 1 separately for SA and VA proteins.<b>S File 6 – Dendrogram with modules</b> Clustering dendrograms of the SA and VA proteins, respectively, with dissimilarity based on topological overlap, together with assigned module colours after the Dynamic tree cut and subsequent merging of highly similar modules (module eigengene correlation &gt; 0.8). The colours were assigned independently for SA and VA dendrogram. <b>S File 7 – SA WGCNA and module-trait relationships results</b> The table contains the module membership and gene significance with the respective p-values of the WGCNA and module-trait relationship analysis for the SA protein expression. <b>S File 8 – VA WGCNA and module-trait relationships results</b> The table contains the module membership and gene significance with the respective p-values of the WGCNA and module-trait relationship analysis for the VA protein expression.<br>**补充材料** **补充文件1——带标准化与插补强度值的数据集** 该数据集源自MaxQuant搜索得到的proteinGroups.txt文件，经过如下处理：a) 移除假阳性命中与污染蛋白质组；b) 剔除4组男性样本对及1组异常样本对；b) 对蛋白质组强度进行以2为底的对数变换；c) 采用LoessF标准化方法；d) 使用imp4p软件包完成缺失值插补；e) 过滤两类样本：VA组或SA组中测量强度值少于8个的蛋白质组，以及在SA或VA样本组中被少于2个肽段鉴定到的蛋白质组。经上述过滤得到的标准化与插补强度数据集，被用于基于Limma R包的比较分析。 **补充文件2——Limma差异表达分析结果** 采用LIMMA R包开展差异表达分析²⁸。用于对比SA与VA样本配对差异的线性模型，通过将批次号作为变量纳入模型以校正批次效应。借助LIMMA包的适配函数，将样本对之间的相关性纳入线性模型²⁹。随后，使用LIMMA包中实现的Benjamini与Hochberg多重假设检验校正方法³⁰对分析结果进行校正。 **补充文件3a——SA组Reactome富集通路分析结果** 将所有显著上调的SA蛋白质的UniProt登录号列表提交至Reactome数据分析工具，得到该文件。 **补充文件3b——VA组Reactome富集通路分析结果** 将所有显著上调的VA蛋白质的UniProt登录号列表提交至Reactome数据分析工具，得到该文件。 **补充文件4a——所有差异表达蛋白质的通路富集分析结果** 使用Cytoscape的ClueGO插件，将上调的SA与VA蛋白质分别作为独立分组，开展Reactome通路与反应富集分析。分析采用默认参数，仅展示p值<0.05的结果。 **补充文件4b——SA组上调蛋白质的通路富集分析结果** 使用Cytoscape的ClueGO插件，针对SA组上调蛋白质开展Reactome通路与反应富集分析。分析采用默认参数，仅展示p值<0.05的结果。 **补充文件4c——VA组上调蛋白质的通路富集分析结果** 使用Cytoscape的ClueGO插件，针对VA组上调蛋白质开展Reactome通路与反应富集分析。分析采用默认参数，仅展示p值<0.05的结果。 **补充文件5——潜在分泌蛋白的SignalP预测结果** 使用SignalP-5.0服务器开展潜在分泌蛋白分析得到的输出结果。本次分析分别针对SA与VA蛋白质，采用以2为底的折叠变化（log₂FC）>1的差异最显著蛋白质的FASTA序列。 **补充文件6——带模块的聚类树状图** 分别基于SA与VA蛋白质的拓扑重叠性构建非相似性聚类树状图，结合动态树切割算法与后续高度相似模块（模块特征基因相关性>0.8）的合并步骤，得到已分配模块颜色的聚类树。SA与VA的聚类树分别独立分配颜色。 **补充文件7——SA组WGCNA（加权基因共表达网络分析）及模块-性状关联分析结果** 该表格包含针对SA蛋白质表达的WGCNA及模块-性状关联分析中的模块隶属度、基因显著性及其对应p值。 **补充文件8——VA组WGCNA（加权基因共表达网络分析）及模块-性状关联分析结果** 该表格包含针对VA蛋白质表达的WGCNA及模块-性状关联分析中的模块隶属度、基因显著性及其对应p值。