STED:Supplement Information and Demo data

STED: Single-cell Topic Modeling & Epigenetic Deconvolution

STED is a Python package designed for the analysis of single-cell and bulk multi-omics data. It leverages semi-supervised topic modeling (Anchored CorEx and LDA) to identify cell-type-specific gene modules in single-cell data and uses these models to deconvolve bulk RNA-seq, ATAC-seq and ChIP-seq data, estimating cell type proportions and predicting cell-type-specific epigenetic signals.

STED：单细胞主题建模与表观遗传反卷积。STED是一款面向单细胞与批量多组学数据分析的Python软件包，其采用半监督主题建模技术（锚定CorEx（Anchored CorEx）与潜在狄利克雷分配（Latent Dirichlet Allocation，LDA）），从单细胞数据中识别细胞类型特异性基因模块，并基于这些模型对批量RNA测序（RNA-seq）、转座酶可及性测序（ATAC-seq）及染色质免疫沉淀测序（ChIP-seq）数据开展反卷积分析，实现细胞类型占比估算与细胞类型特异性表观遗传信号预测。