812 simulated expression datasets for differential co-expression analysis

Simulated expression data with knock-outs Description A dataset containing simulated expression dataset. Data is simulated using a dynamical systems model from a network sampled from the S. Cerevisiae regulatory network. The dataset is a list containing the results from the simulation, and other information generated subsequently. Format A named list with 14 elements: simitra numeric, indicating the iteration of the simulation (a total of 1000 were performed and 812 converged) scoresan S4 Matrix, containing vectorised inference scores of applying the methods implemented in the package. These are precomputed predictions inputmodelsa named list, storing the parameters used to sample the initial values of input genes. Proportions, means and variances of each gene is stored for each gene staticnetan igraph object, storing the initial regulatory network (150 node network) infnetan igraph object, representing the true differential network as determined using sensitivity analysis of the model netlayouta matrix (150 x 2), storing the (x, y) positions of nodes for laying out the graph infdensa numeric, network density of the true differential association network numinputa numeric, the number of input genes in the regulatory network. These are genes that have no regulators therefore need to be pre-defined numbimodala numeric, the number of input genes that are knocked-down therefore have a bimodal distribution numtfsa numeric, the number of genes in the network that regulate any other gene (are TFs) numcotargetsa numeric, the number of genes that are co-regulated, i.e. regulated by more than one TF dataan S4 Matrix, the expression data with samples along the columns and genes along the rows. Condition classification (KD vs WT) are stored as attributes of this object tripletsa data frame, consisting of gene triplets representing TF- Target associations conditioned on the gene knocked-down. Triplets are annotated for being in either the direct, influence and association networks sensmatan S4 Matrix, sensitivities of genes to TFs based on perturbation analysis of the simulation modelLoadThis dataset is in the form of an R RDS object. To load it, type the command below in an R console:simdata = readRDS("sim812.rds")带敲除处理的模拟表达数据集（Simulated expression data with knock-outs） ## 数据集描述 本数据集为模拟表达数据集，其数据基于从酿酒酵母（Saccharomyces cerevisiae, S. Cerevisiae）调控网络中采样得到的网络，通过动力学系统模型生成。数据集为包含模拟结果及后续生成的其他信息的列表。 ## 数据格式 该数据集为包含14个命名元素的命名列表，各元素说明如下： 1. `simitra`：数值型变量，代表模拟迭代次数（共执行1000次迭代，其中812次收敛） 2. `scoresan`：S4矩阵（S4 Matrix），存储本包中实现的方法经向量化处理后的推断得分，即预计算得到的预测结果 3. `inputmodels`：命名列表，存储用于采样输入基因初始值的参数，其中为每个基因保存了其比例、均值与方差 4. `staticnet`：igraph对象（igraph object），存储初始调控网络（含150个节点） 5. `infnet`：igraph对象（igraph object），代表通过模型敏感性分析确定的真实差异网络 6. `netlayout`：150×2的矩阵，存储用于绘图布局的节点(x,y)坐标 7. `infdensa`：数值型变量，代表真实差异关联网络的网络密度 8. `numinput`：数值型变量，代表调控网络中的输入基因数量，此类基因无调控因子，需预先指定 9. `numbimodal`：数值型变量，代表被敲低（Knock-Down, KD）的输入基因数量，此类基因的表达分布呈双峰态 10. `numtfs`：数值型变量，代表网络中可调控其他基因的基因数量（Transcription Factor, TF，即转录因子） 11. `numcotargets`：数值型变量，代表共调控基因的数量，即被多个转录因子共同调控的基因 12. `data`：S4矩阵（S4 Matrix），存储表达数据，其中列代表样本、行代表基因，样本的分组信息（敲低组KD vs 野生型Wild-Type, WT）以该矩阵的属性形式保存 13. `triplets`：数据框，包含基因三元组，用于代表以敲低基因为条件的转录因子-靶标关联关系，三元组会被标注其所属的网络类型：直接网络、影响网络或关联网络 14. `sensmat`：S4矩阵（S4 Matrix），存储基于模拟模型扰动分析得到的基因对转录因子的敏感性得分 ## 加载方式 本数据集以R语言RDS格式存储。若需加载该数据集，请在R控制台中执行如下命令：`simdata = readRDS("sim812.rds")`