Table 1.

Summary of reviewed methods.

Method	Category	Language	Software	Reference	Released Date	Advantages	Disadvantages	Use histology image
HMRF	Spatial clustering	R; Python; C	https://bitbucket.org/qzhud- fci/smfishhmrf-py	Identification of spatially associated subpopulations by combining scRNAseq and sequential fluorescence in situ hybridization data [37]	2018-10-29	Can simultaneously detect the combinatorial pattern of all profiled genes.	The classification of a small number of isolated cells as domains may be questionable. Cannot incorporate histology information in its model.	No
SpaCell	Spatial clustering	Python	https://github.com/BiomedicalMachineLearning/SpaCell	SpaCell: integrating tissue morphology and spatial gene expression to predict disease cells [41]	2020-04-01	Can combine histology image data and spatial gene expression data for joint clustering. Can automatically and quantitatively identify cell types and disease stages.	Spot location information is not utilized in the model.	Yes
stLearn	Spatial clustering	Python	https://github.com/BiomedicalMachineLearning/stLearn	stLearn: integrating spatial location, tissue morphology and gene expression to find cell types, cell–cell interactions and spatial trajectories within undissociated tissues [43]	2020-05-31	Can integrate gene expression and spatial distance information and histology image information. Applicable to any SRT data as long as tissue morphology, spatial location, and gene expression information are simultaneously captured.	Cannot be applied to data without histology images.	Yes
BayesSpace	Spatial clustering; enhancement of gene expression resolution	R; C++	https://github.com/edward130603/BayesSpace	BayesSpace enables the robust characterization of spatial gene expression architecture in tissue sections at increased resolution [45]	2020-09-05	Account for spatial dependency in clustering analysis. Can generate enhanced resolution gene expression data.	Cannot incorporate histology information.	No
SpaGCN	Spatial clustering; identification of spatially variable genes	Python	https://github.com/jianhuupenn/SpaGCN	Integrating gene expression, spatial location and histology to identify spatial domains and spatially variable genes by graph convolutional network [36]	2020–11-30	Jointly consider spatial domain identification and SVG detection. Can integrate gene expression, spatial location and histology information (when available) in spatial domain identification.Computationally fast and memory efficient.	Cannot account for cell type variations in spatially variable gene detection.	Yes
Trendsceek	Identification of spatially variable genes	R	https://github.com/edsgard/trendsceek	Identification of spatial expression trends in single-cell gene expression data [46]	2018–03-19	Perform a gene-level test that incorporates both spatial and expression-level information.	Cannot account for cell type variations in spatially variable gene detection.	No
SpatialDE	Identification of spatially variable genes	Python	https://github.com/Teichlab/SpatialDE	SpatialDE: identification of spatially variable genes [47]	2018–03-19	Use of a principled statistical approach to model spatial dependency of gene expression.	Rely on asymptotic normality and minimal P-value-combination rules for hypothesis testing, which may lead to false positives and loss of power. Cannot account for cell type variations in spatially variable gene detection.	No
SPARK	Identification of spatially variable genes	R; C++	https://github.com/xzhoulab/SPARK	Statistical analysis of spatial expression patterns for spatially resolved transcriptomic studies [48]	2020–01-27	Explicit modeling of gene expression as count data. Use of kernel approaches to model spatial dependency of gene expression.	Computationally slow and memory consuming. Cannot account for cell type variations in spatially variable gene detection.	No
Stereoscope	Cell-type deconvolution	Python	https://github.com/almaan/stereoscope	Spatial mapping of cell types by integration of transcriptomics data [56]	2019–12-13	First method for cell-type deconvolution in SRT.	Assume both SRT and scRNA-seq data follow a negative binomial distribution. Do not account for spatial dependency of gene expression.	No
RCTD	Cell-type deconvolution; enhancement of gene expression resolution	R	https://github.com/dmcable/RCTD	Robust decomposition of cell type mixtures in spatial transcriptomics [57]	2020–05-08	Can correct for platform differences between SRT data and scRNA-seq reference. Can restrict deconvolution only to the most likely cell types.	Do not explicitly model spatial dependency of gene expression. Assume platform effects are shared among all cell types.	No
SPOTlight	Cell-type deconvolution	R	https://github.com/MarcElosua/SPOTlight_deconvolution_analysis	SPOTlight:Seeded NMF regression to Deconvolute Spatial Transcriptomics Spots with Single-Cell Transcriptomes [58]	2020–06-04	A small number of cells per cell-type is sufficient to train the model.	Need prior information of cell-type-specific marker genes.	No
Cell2location	Cell-type deconvolution	Python	https://github.com/BayraktarLab/cell2location	Comprehensive mapping of tissue cell architecture via integrated single cell and spatial transcriptomics [59]	2020–11-15	Can accurately infer the presence of rare cell types. Can provide estimates of relative cell type fractions along with additionally estimates of absolute cell type abundance.	Do not explicitly model spatial dependency of gene expression. Require the user to define hyperparameters for the average number of cells, cell types, and tissue zones per spot and the average difference in technical sensitivity between scRNA-seq and SRT data.	No
spatialDWLS	Cell-type deconvolution	R	https://github.com/RubD/Giotto	SpatialDWLS: accurate deconvolution of spatial transcriptomic data [60]	2021–05-10	Can restrict deconvolution only to the most likely cell types.	Do not account for spatial dependency of gene expression in deconvolution.	No
XFuse	Enhancement of gene expression resolution	Python	https://github.com/ludvb/xfuse	Super-resolved spatial transcriptomics by deep data fusion [62]	2020–03-13	Can infer spatial gene expression at the same resolution as the histology image data.	Assume the gene expression and histology image share the same latent state.	Yes
Giotto	Cell-cell communications	R	https://github.com/RubD/Giotto	Giotto, a pipeline for integrative analysis and visualization of single-cell spatial transcriptomic data [61]	2020-05-30	Can identify genes whose expression variation within a cell type is significantly associated with an interacting cell type.	Only focus on unsupervised correlation-based analysis, thus may fail to identify interactions that are limited to a specific area, specific cell types, or that are related to more complex patterns.	No
SpaOTsc	Cell-cell communications	Python	https://github.com/zcang/SpaOTsc	Inferring spatial and signaling relationships between cells from single cell transcriptomic data [69]	2020–04-29	Model both direct and indirect cell–cell communications.	The computation of the cell–cell distance inference can become intractable when the dataset is excessively large [69]. HYPERLINK "SPS:refid::bib69"	No
MISTy	Cell-cell communications	R	https://github.com/saezlab/mistyR	Explainable multi-view framework for dissecting inter-cellular signaling from highly multiplexed spatial data [71]	2020–05-10	Can build multiple views focusing on different spatial or functional contexts to dissect different effects.	Rely on a radius parameter to determine the number of cells to be included in each view.	No