yuwang-vumc / stimage

R package for integration analysis of spatial transcriptomics and morphology images

stImage

StImage is a R package for integrated analysis of spatial transcriptomics and the corresponding modality images. StImage is freely available at https://github.com/YuWang-VUMC/stImage.

Installation

Before installing stImage, dependencies should be installed first:

# SPARK/SPARKX and SpatialPCA
library(devtools)
install_github('xzhoulab/SPARK')
install_github("shangll123/SpatialPCA")

#BioConductor packages
if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

BiocManager::install(c("rhdf5", "SingleCellExperiment", "BayesSpace", "omicade4"))

Users should also make sure to successfully install the tensorflow and keras following instruction Tensorflow for R.

#installation of tensorflow
install.packages("tensorflow")

#create python virtualenv. If already installed python, replace 'install_python()' with the path of excutable python.
library(reticulate)
path_to_python <- install_python()
virtualenv_create("r-reticulate", python = path_to_python)

library(tensorflow)
install_tensorflow(envname = "r-reticulate")

#installation of keras
install.packages("keras")
library(keras)
install_keras(envname = "r-reticulate")

#testing the installation
library(tensorflow)
tf$constant("Hello Tensorflow!")

Once Tensorflow and keras were successfully installed, you can then install the latest version of stImage from GitHub with:

library(devtools)
install_github("YuWang-VUMC/stImage")

Tutorial

• Adult Mouse Olfactory Bulb Visium Dataset processed by Space Ranger 2.0.0

• HER2 breast cancer data by ST platform

The tutorial includes main example codes for multiple spatial transcriptomics datasets (e.g. Adult Mouse Olfactory Bulb and Human breast tumor)

License

stImage is licensed under the MIT License.

Citation

stImage xxx.

doi: xxx