、

scPOEM: scRNA-seq and scATAC-seq Paired Embedding Database

Overview: This database presents an integrated analysis of paired scRNA-seq and scATAC-seq data by scPOEM, which generates a joint low-dimensional embedding of genes and peaks for each cell type. Specifically, after quality control and the selection of 3,000 highly variable genes (HVGs), heterogeneous regulatory networks, including gene–gene, peak–peak, and gene–peak interactions, were inferred using LASSO, random forest, and XGBoost, and are then jointly embedded into a low-dimensional latent space using the scPOEM R package.
R Package: scPOEM GitHub: scPOEM
References:
1. Controlled noise: evidence of epigenetic regulation of single-cell expression variability.
DOI: 10.1093/bioinformatics/btae457

2. scPOEM: robust co-embedding of peaks and genes revealing peak–gene regulation
DOI: 10.1093/bioinformatics/btaf483
Database Developer:
Xinyue Zheng, Chenge Gao, Xixian Zhou, Ziyu Liu, Yeran Chen, Youshi Chang,
Yan Zhong(From East China Normal University, https://faculty.ecnu.edu.cn/_s35/zy2_en_21397/main.psp)
Showing 13 items
Data Name Source Detailed Information Description
10xPBMC 10xPBMC Detailed Information This dataset contains 10,000 peripheral blood mononuclear cells (PBMC) from a healthy 25-year-old female donor (AllCells), with granulocytes removed by cell sorting and nuclei isolated, profiled using 10x Genomics.
sciCAR NCBI GEO Detailed Information Single-cell RNA-seq and ATAC-seq data generated by sci-CAR, profiling chromatin accessibility and gene expression in HEK293T, NIH/3T3, A549 cells under DEX treatment (0, 1, 3 hours) and wild-type mouse kidney cells.
10xNeuron NCBI GEO Detailed Information Single-cell RNA-seq and single-nucleus ATAC-seq of juvenile mouse small intestine enteric neurons and 3D neurosphere cultures. It identifies multiple transcriptional subpopulations of enteric glial cells.
SNAREseq NCBI GEO Detailed Information This dataset uses SNARE-seq to profile 1,047 nuclei from a human cell line mixture.
SNAREseq-human NEMO Archive Detailed Information This dataset uses SNARE-seq to profile 5,081 nuclei from neonatal.
SNAREseq-marmoset NEMO Archive Detailed Information This dataset uses SNARE-seq to profile 10,309 nuclei from adult mouse brains.
SNAREseq-mouse NCBI GEO Detailed Information This dataset applies SNARE-seq to adult mouse tissues (skin, brain, lung) to jointly profile chromatin accessibility and gene expression in the same cells from cell lines and mouse tissues.
human-brain Figshare Detailed Information This dataset profiles single-cell chromatin accessibility and gene expression from four mid-gestation human cortical samples across 8 weeks of development.
human-lung NCBI GEO Detailed Information Nuclei were extracted directly from frozen tissue to generate a human lung single-cell multi-omics dataset, including RNA-based cell type annotation and peak-by-cell chromatin accessibility matrices. For each of the 54 cell types, the top 10,000 cell type–specific accessible sites were retained.
human-pancreas NCBI GEO Detailed Information This dataset includes single-cell RNA matrices for cell type annotation and peak-by-cell chromatin accessibility matrices for downstream analyses, with up to 800 randomly sampled cells per cell type per tissue.
human-stomach NCBI GEO Detailed Information Human gastric cells were analyzed, including single-cell gene-by-cell matrices for cell type annotation and single-cell peak-by-cell matrices for downstream analyses, generated for each tissue. For each cell type, up to 800 cells were randomly sampled; if fewer than 800 cells of a given cell type were available in a tissue, all cells were retained.
human-thymus NCBI GEO Detailed Information Human thymus single-cell data include gene-by-cell matrices for cell type annotation and peak-by-cell chromatin accessibility matrices for downstream analyses. For each cell type, up to 800 cells were randomly sampled per tissue, with all cells retained when fewer were available.
multiome-object Zenodo Detailed Information A mixed population of iPS and K562 cells and retinal organoids at 18 and 24 weeks, processed with the Phospho-seq Multi protocol, were profiled by whole-cell protein analysis and scATAC-seq.