STIE

Spatial Transcriptome Image and Expression integration enables single-cell level spatial transcriptomics data analysis

Description

STIE is a novel computational method tailored for spatial transcriptomics data analysis, which integrated spot level gene expression, nuclear segmentation, and nuclear morphology to perform cell type deconvolution/convolution and clustering, therefore enabling the single-cell level spatial transcriptomics anlayiss.

Dependencies on packages

STIE has been tested on GNU/Linux but should run on all major operating systems. STIE depends on the following packages:

• imagemagick - imagemagick (>=7.1.0)
• ImageJ/Fiji - ImageJ/Fiji
• R package: magick; EBImage; Seurat; CellChat; quadprog;

Installation

1) Obtain R (>=3.6)

Clear instructions for different version can be found here: http://cran.fhcrc.org/

2) Install the dependent R packages

# install R packages of computing
> install.packages(c("quadprog"))

# install magick
> install.packages("magick")

# install EBImage
> if (!require("BiocManager", quietly = TRUE))
>     install.packages("BiocManager")
> BiocManager::install("EBImage")

# install CellChat
> BiocManager::install("ComplexHeatmap")
> devtools::install_github("sqjin/CellChat")

# install Seurat
> install.packages('Seurat')

3) Install the STIE R package

git clone https://github.com/zhushijia/STIE.git
R CMD INSTALL -l userFolder STIE

Tutorial

See our Wiki (long); Vignette (short); and NucleusSegmentation

Citation

STIE: Single-cell level deconvolution, convolution, and clustering in spatial transcriptomics by aligning spot level transcriptome to nuclear morphology (preprint)