Table 3:

Overview of key analytical tools for single-cell epigenomic data

Method	Description	Exemplary studies with application in human tissues
Data processing and clustering
ArchR136	Iterative latent semantic indexing using read counts in genomic windows followed by cluster-specific peak calls	Clustering of 70k nuclei from 10x scATAC-seq of isocortex, striatum, hippocampus, and substantia nigra181.
Signac156	Latent semantic indexing using read counts in peak calls	Clustering of 27k nuclei from kidney 10x scATAC-seq184, 81k nuclei from motor cortex SNARE-seq2182, 44k nuclei from pituitary gland 10x scATAC-seq186.
Cusanovich et al60	Latent semantic indexing using read counts in genomic windows and cluster-specific peak calls	Clustering of 1.6k nuclei from pancreatic islet sciATAC-seq187, 35k nuclei from bone marrow and blood 10x scATAC-seq194, 63.8k nuclei from bone marrow and blood 10x scATAC-seq58, 791k nuclei from sciATAC-seq3 of fetal tissue63, 31k nuclei from fetal cortex 10x scATAC-seq191.
SnapATAC137	Spectral embedding using Jaccard similarity of read counts in genomic windows	Clustering of 79.5k nuclei from heart snATAC-seq195, 616k nuclei from snATAC-seq of 30 tissues30, 12.7k nuclei from kidney 10x scATAC-seq198, 12.5k nuclei from frontal cortex snATAC-seq48
Scanpy202/EpiScanpy149	Principle components analysis of read counts in genomic windows, peak calls, or other features	Clustering of 91k nuclei from lung snATAC-seq188, 15k nuclei from pancreatic islet snATAC-seq199, 131.5k nuclei from islet and PBMCs snATAC-seq and 10x scATAC-seq196.
cisTopic138	Latent Dirichlet allocation of read counts in peak calls	-
SCALE139	Variational autoencoder of read counts in peak calls	-
BROCKMAN151	Principle components analysis of sequence k-mer counts	-
Scasat153	Multidimensional scaling using Jaccard similarity of read counts in peak calls	-
Cell Ranger ATAC58	Latent semantic analysis of read counts in peak calls	Clustering of 10x scATAC-seq of 7k nuclei from atherosclerotic lesions197.
AllCools	Consensus clustering using Leiden algorithm on principle components of DNA methylation levels in genomic windows.
Data integration
Seurat integration (v3)156	Diagonal and horizontal integration using canonical correlation analysis	Diagonal integration of 130k nuclei from prefrontal cortex snATAC-seq with snRNA-seq192, 7k nuclei of 10x scATAC-seq from atherosclerotic lesions with snRNA-seq197, 31k nuclei from fetal cortex 10x scATAC-seq with snRNA-seq191, 81k nuclei from motor cortex SNARE-seq2182, 79.5k nuclei from heart snATAC-seq with snRNA-seq195, skeletal muscle 10x scATAC-seq, 10x scMultiome with snRNA-seq200, bone marrow and blood 10x scATAC-seq with snRNA-seq194.
Liger166	Diagonal and horizontal integration using integrative non-negative matrix factorization	Diagonal integration of skeletal muscle 10x scATAC-seq, 10x scMultiome with snRNA-seq200
SingleCellFusion48	Diagonal and horizontal integration using mutual nearest neighbors	Diagonal integration of frontal cortex snmCAT-seq and snATAC-seq48.
SnapATAC137	Horizonal integration using landmark diffusion maps	Horizontal integration of 79.5k nuclei from heart snATAC-seq195, 616k nuclei from sci-ATAC-seq of 30 tissues30.
EpiScanpy149	Horizonal integration using batch-corrected k-nearest neighbors	-
ArchR136	Horizontal integration using estimated latent semantic indexing	-
MNN158	Horizontal integration using mutual nearest neighbors	Horizontal integration for batch correction of prefrontal cortex 10x scATAC-seq192, snATAC-seq of 30 tissues30, heart snATAC-seq195.
Harmony157	Horizontal integration using linear mixed model correction	Horizontal integration for batch correction of pancreatic islet snATAC-seq199, pancreas and PBMCs 10x scATAC-seq and snATAC-seq196, lung snATAC-seq188, isocortex, striatum, hippocampus and substantia nigra 10x scATAC-seq181, sci-ATAC-seq3 of 15 fetal tissues63, kidney 10x scATAC-seq184.
Symphony159	Horizontal integration for reference mapping using linear mixed model correction	-
Seurat WNN (v4)161	Vertical integration using weighted nearest neighbors	Vertical integration of 15k nuclei from 10x scMultiome in pituitary gland186.
MOFA+162	Vertical integration using stochastic variational inference	-
Downstream analysis
Cicero175	Co-accessibility between peak calls and cis-co-accessibility networks using graphical LASSO	Analysis of co-accessible sites in pancreatic islet snATAC-seq199, pancreas and PBMCs 10x scATAC-seq and snATAC-seq196, lung snATAC-seq188, heart snATAC-seq195, skeletal muscle 10x scATAC-seq200, prefrontal cortex 10x scATAC-seq192, atherosclerotic lesion 10x scATAC-seq197, 10x scATAC-seq of isocortex, striatum, hippocampus, and substantia nigra181, kidney snATAC-seq184, bone marrow and peripheral blood snATAC-seq58.
ChromVAR174	Sequence motif enrichment using bias-corrected deviations	Sequence motif enrichments of cells in pancreatic islet snATAC-seq199, pancreas and PBMCs 10x scATAC-seq and snATAC-seq196, atherosclerotic lesion 10x scATAC-seq197, kidney 10x scATAC-seq184, prefrontal cortex 10x scATAC-seq192, heart snATAC-seq195, fetal cortex 10x scATAC-seq191, bone marrow and peripheral blood 10x scATAC-seq58.
Monocle (v2)177	Pseudo-time trajectory ordering using reverse graph embedding	Pseudo-time ordering of cells in 10x scATAC-seq of atherosclerotic lesions197.
Monocle (v3)178	Pseudo-time trajectory ordering using partitioned approximate graph abstraction	Pseudo-time ordering of cells sites in pancreatic islet snATAC-seq199, prefrontal cortex 10x scATAC-seq191,192, sci-ATAC-seq3 of 15 fetal tissues63, kidney 10x scATAC-seq184.
Slingshot179	Pseudo-time trajectory ordering using simultaneous principal curves	-
Destiny180	Pseudo-time trajectory ordering using diffusion maps	-
Signac135	Multiple methods including peak calling, cluster-specific differential peaks, peak-to-gene links, and transcription factor footprinting	Cluster-specific differential peaks in 10x scATAC-seq of atherosclerotic lesions197, kidney 10x scATAC-seq184, Peak-to-gene links in skeletal muscle 10x scMultiome200.
ArchR136	Multiple methods including peak calling, cluster-specific differential peaks, gene activity, peak-to-gene links, transcription factor footprinting, and pseudotime trajectory ordering	Cluster-specific analyses of isocortex, striatum, hippocampus, and substantia nigra 10x scATAC-seq181, transcription factor footprinting and peak-based analysis in prefrontal cortex 10x scATAC-seq192.
SnapATAC137	Multiple methods including peak calling, cluster-specific differential peaks, and peak-to-gene links	Cluster-specific differential peaks in kidney 10x scATAC-seq198.
AllCools	Differential methylated regions and genes

ArchR, single-cell analysis of regulatory chromatin in R, SnapATAC, single nucleus analysis pipeline for ATAC-seq, Scanpy, single cell analysis in python; EpiScanpy, epigenomics single cell analysis in python; cisTopic, cis-regulatory topic modeling on single-cell ATAC-seq data; SCALE, single-cell ATAC-seq analysis via latent feature extraction; BROCKMAN, Brockman representation of chromatin by K-mers in mark-associated nucleotides; Scasat, single cell ATAC-seq analysis tool; LIGER, linked Inference of genomic experimental relationships; MNN, mutual nearest neighbour, WNN, weighted nearest neighbor, MOFA+, multi-omics factor analysis