Table 2.

Overview of analytical features of single-cell methylome analysis tools

	General information		Input data			Data processing					Cell state annotation			Modeling		Multi-omics			Data presentation
Name	Programming language	Benchmarked dataset	Epigenome	scRNA-seq integration	Data format	Filtering	Normalization	Imputation	Feature matrix	Feature engineering	Motif analysis	Cell typing	Cell trajectory	Clustering	Differential methylation	Label transfer	Gene activity scoring	Cell space	Visualization	Ref.
BPRmeth	R	scNMT	M		MC	F+V		+	RC	SVLR				+	+		+		+	[134]
DeepCpG	P	scBS-seq scRRBS	M		MC			+	RC	CNN	+			+	+				+	[139]
MELISSA	R	scM&T-seq scBS-seq	M		MC		+	+	RC	BI				+					+	[143]
Epiclomal	P+R	scBS-seq scRRBS scWGBS	M		MC	F+V	+	+	RC	BI				+						[144]
MAPLE	R	snmC-seq scM&T-seq scNMT-seq	M	+	AB					ENS						S	+	DS*		[211]
MethylStar	P+R+Sh	scBS-seq	M		FQ			MI												[141]
scMET	R	snmC-seq scNMT-seq	M		MC			+	RC	BI					+					[189]
EpiScanpy	P	snmC-seq	M/A		MC	F+V+D	+	+	RC	NG			+	+	+		A		+	[138]
coupleCoC+	ML	snmC-seq	M/A	+	MG			+	GC	ITC				+			A	DS	+	[240]
ALLCools	P	snmC-seq2	M/A	+	PY	F+V+D	+		RC	+	+	+		+	+			DS	+	[66]
MATCHER	P	scM&T-seq scGEM	M/A	+	MB								+	+	+	+		DS		[218]
LIGER	R	snmC-seq	M/A	+	MG				FC	MF		+		+	+	+		DS	+	[210]
scAI	R	scM&T-seq	M/A	+	MB	F	+		FC	MF				+	+			CA	+	[228]
MOFA+	R	scM&T-seq	M/A	+	MO		+		FC	MF				+	+			CA	+	[227]

Note: The epigenetic input of these tools is depicted by M if they are designed solely for single-cell methylation-seq or M/A if the input can be either single-cell methylation-seq or single-cell ATAC-seq. ‘+’ indicates that the feature is supported or the functionality is through another specified software. For convenience, we have included the methylation datasets used by the original papers of these tools, for testing and training purposes; single-cell datasets not containing methylation have been omitted (scRNA-seq, scATAC-seq, or both). scRNA-seq, single-cell RNA sequencing; scATAC-seq, single-cell assay for transposase-accessible chromatin sequencing; P, Python; Sh, Shell/BASH; ML, MATLAB; M, DNA methylation; A, chromatin accessibility; MC, methylation call; AB, aligned read in BAM; FQ, FASTQ raw read; MG, methylation call by gene; PY, processed read in the YAPS MCDS format; MB, binarized methylation call; MO, MultiAssayExperiment Object; F, filtering CpGs and cells by sequencing depth or data sparsity; V, filtering CpGs by methylation variation; D, doublet detection and filtering; MI, conducted through METHImpute; RC, genomic region-by-cell matrix; GC, gene-by-cell matrix; FC, factor-by-cell matrix (factor includes shared factors and non-shared factors); SVLR, support vector linear regression; CNN, convolutional neural network; BI, Bayesian inference; ENS, ensemble machine learning (CNN + elastic net + random forest); NG, neighbor graph and graph-based clustering (Louvain, Leiden, etc.); ITC, information-theoretic co-clustering; MF, matrix factorization; S, conducted through Seurat; CA, integration of co-assays; DS, integration of data from different sample spaces; DS*, integration of data from different sample spaces but using co-assay data for regressor training.