Additional file 2 of vmrseq: probabilistic modeling of single-cell methylation heterogeneity

Additional file 2: Table S1-S9. Table S1: Performance of clustering cells in Luo et al. [12] data with VMRs overlapped with targeted features. Default parameters, g = 100 and $$\theta = 0.7$$ θ = 0.7 , were used for calculation of the nearest neighbor score. The ‘broad classes’ refer to the 2 broad neuronal classesin Luo et al. [12] data, and ‘subtypes’ refer to 15 subtypes within those two classes. Gene promoters were defined as $$pm 2$$ p m 2 -kb windows around the transcription start sites of genes extracted from ENSEMBL version 80 [55]. Additional File 2: Table S9 lists data sources of the histone ChIP-seq annotation information. Table S2: Table with sample metadata from the mouse frontal cortex dataset. Table S3: Table with sample metadata from the multi-omic mouse gastrulation dataset. Table S4: Table with the correlation between gene expression and VMR/promoter methylation. Table S5: Table with the subtypes included in training parameters in transition probability and emission probability used as default in this study. Table S6: Table with the trained transition probability. Table S7: Table with the trained beta prior parameters of the emission probability for methylated grouping. Table S8: Table with the trained beta prior parameters of the emission probability for methylated grouping. Table S9: Table with data sources of histone ChIP-seq annotation information used in the experiments.

补充文件2：表S1至S9。
表S1：基于Luo等人[12]数据中与靶向特征重叠的可变甲基化区域（Variable Methylation Region, VMR）进行细胞聚类的性能表现。本研究采用默认参数（g=100，θ=0.7）计算最近邻分数。其中“大类”指Luo等人[12]数据中的2个宽泛神经元类别，“亚型”则指这两类下的15个细胞亚型。基因启动子定义为从ENSEMBL版本80[55]中提取的基因转录起始位点上下游±2 kb的窗口区域。
补充文件2中的表S9列出了组蛋白染色质免疫沉淀测序（Chromatin Immunoprecipitation Sequencing, ChIP-seq）注释信息的数据源。
表S2：小鼠前额叶皮层数据集的样本元数据表。
表S3：多组学小鼠原肠胚形成数据集的样本元数据表。
表S4：基因表达与VMR/启动子甲基化水平的相关性表。
表S5：列有本研究默认采用的转移概率与发射概率训练参数所涉及的细胞亚型。
表S6：已训练得到的转移概率表。
表S7：甲基化分组发射概率的β先验参数训练结果表。
表S8：甲基化分组发射概率的β先验参数训练结果表。
表S9：本研究实验中所用组蛋白ChIP-seq注释信息的数据源表。