Multi-omics data analysis for rare population inference using single-cell graph transformer

## MarsGT: For rare cell identification from matched scRNA-seq (snRNA-seq) and scATAC-seq (snATAC-seq),includes genes, enhancers, and cells in a heterogeneous graph to simultaneously identify major cell clusters and rare cell clusters based on eRegulon. ## Data Collection The data was collected using GEO Database. ## Data Format The data is stored as TSV file and MTX file where each row represents a gene and each column represents a sample. ## Variables - Gene IDs: Gene Symbols (e.g., MALAT1) - Sample IDs: Sample identifiers (e.g., AAACATGCAAATTCGT-1) - Expression level: Row gene expression level.# MarsGT：一款用于匹配的单细胞RNA测序（scRNA-seq）与单核RNA测序（snRNA-seq）、单细胞转座酶可及性测序（scATAC-seq）与单核转座酶可及性测序（snATAC-seq）数据的稀有细胞识别工具。该方法将基因、增强子与细胞构建为异质图，基于增强子调控单元（eRegulon）同时识别主要细胞簇与稀有细胞簇。 ## 数据集采集 本数据集采集自GEO数据库（Gene Expression Omnibus）。 ## 数据格式 数据集以TSV文件与MTX文件格式存储，其中每行代表一个基因，每列代表一个样本。 ## 变量说明 - 基因标识：基因符号（例如MALAT1） - 样本标识：样本标识符（例如AAACATGCAAATTCGT-1） - 表达水平：对应行基因的表达量