ST-Assign

This directory contains the source code required to replicate analyses presented in the manuscript:

A. Geras, K. Domżał, E. Szczurek, Joint cell type identification in spatial transcriptomics and single-cell RNA sequencing data, in review.

Running ST-Assign

ST-Assign can be run from the code directory as follows:

bash  ST-Assign-run.sh 'input_data' 'results' 

The arguments of the bash script:

• input_data - the directory containing input data,
• results - the directory dedicated to results.

Expected input

ST-Assign takes as input the following files provided in the input_data directory:

• param.txt - a file containing run setting (hyperparameters, number of iterations, etc.),
• matB.csv - a binary matrix with prior knowledge about marker genes,
• C_gs.csv - ST gene expression data,
• C_gs.csv - scRNA-seq gene expression data,
• n_cells.csv - estimates for the number of cells in each ST spot,
• rho.csv - cross-platform factor computed based on gene expression data.

Note: The order of genes in C_gs.csv and C_gs.csv should match the order of genes in matB.csv. Moreover, the order of spots in the file C_gs.csv should match the order in the matB.csv file.

Refer to the Expected Input and Output file in this repository for more information about input formats.

ST-Assing output

• est_M.csv - cell type mixture decomposition in ST spots (size: number of spots vs. number of cell types),
• res_TC.csv - cell type annotations of single cells (size: trajectories after burn in times number of cells).

Refer to the Expected Input and Output file in this repository for more information about input and output formats.

Test example

The test-example directory contains exemplificatory data. After running ST-Assign on this dataset, results can be visualized using the Visualisation.rmd file.

Packages

ST-Assign was developed under:

• python version 3.9.18
• tensorflow version 2.13.0,
• tensorflow-probability version 0.21.0,
• numpy version 1.24.3,
• pandas version 1.5.2,
• scipy version 1.10.0.