deconvATAC

The deconvATAC package provides code used in our benchmarking study for deconvoluting spatialATAC data via deconvolution tools designed for spatial transcriptomics. In our study, we benchmark five top-performing spatial transcriptomics deconvolution methods. deconvATAC additionally provides a framework for simulating spatial multi-modal data from dissociated single-cell data, as well as metrics for evaluating the performance of deconvolution.

Please refer to the documentation.

Data used in this study is available on Zenodo

Installation

Create conda environment

conda create -n deconvATAC python=3.9 r-base=4.3.0
conda activate deconvATAC

Installing deconvATAC

First, clone the directory:

git clone https://github.com/theislab/deconvATAC.git

Install the package:

cd deconvATAC
pip install .

Installing optional dependencies

You can install the dependencies needed for the python-based deconvolution methods with:

pip install .[cell2location] # note: for zsh shell, please use brackets: '.[cell2location]'
pip install .[tangram]
pip install .[destvi]

RCTD

For installing RCTD, please use the following

conda install bioconda::r-spacexr

In your R terminal, install

if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")
BiocManager::install("S4Vectors")
BiocManager::install("SingleCellExperiment")

SpatialDWLS

For SpatialDWLS, the Giotto package needs to be installed. Please follow the installation guidelines in the Giotto documentation for installation of the package.

Citation

t.b.a