SDePER (Spatial Deconvolution method with Platform Effect Removal) is a hybrid machine learning and regression method to deconvolve Spatial barcoding-based transcriptomic data using reference single-cell RNA sequencing data, considering platform effects removal, sparsity of cell types per capture spot and across-spots spatial correlation in cell type compositions. SDePER is also able to impute cell type compositions and gene expression at unmeasured locations in a tissue map with enhanced resolution.

SDePER currently supports only Linux operating systems such as Ubuntu, and is compatible with Python versions 3.9.12 up to but not including 3.11.

SDePER can be installed via conda

conda create -n sdeper-env -c bioconda -c conda-forge python=3.9.12 sdeper

or pip

conda create -n sdeper-env python=3.9.12
conda activate sdeper-env
pip install sdeper

SDePER requires 4 input files for cell type deconvolution:

1. raw nUMI counts of spatial transcriptomics data (spots × genes): spatial.csv
2. raw nUMI counts of reference scRNA-seq data (cells × genes): scrna_ref.csv
3. cell type annotations for all cells in scRNA-seq data (cells × 1): scrna_anno.csv
4. adjacency matrix of spots in spatial transcriptomics data (spots × spots): adjacency.csv

To start cell type deconvolution by running

runDeconvolution -q spatial.csv -r scrna_ref.csv -c scrna_anno.csv -a adjacency.csv

Homepage: https://az7jh2.github.io/SDePER/.

Full Documentation for SDePER is available on Read the Docs.

Example data and Analysis using SDePER are available in our GitHub repository SDePER_Analysis.