linboqiao / GTAD

GTAD: A Graph-based Approach for Cell Spatial Composition Inference from Integrated scRNA-seq and ST-seq Data

Requirements

python

• tensorflow=2.12.0
• scanpy=1.9.3
• numpy =1.23.5
• pandas=2.0.2
• scikit-learn=1.0.2
• scipy 1.11.1

R

• R=4.2.0
• Seurat=4.3.0
• DropletUtils=1.18.1

Run the model

This model requires you to create a 'data' folder in the current directory to store the data for both scRNA-seq and ST.

• For scRNA-seq, you will need 'scRNA_data.csv' and 'scRNA_meta.csv', which respectively represent the gene expression matrix of its cells and metadata. The gene expression matrix should be in the 'gene×cell' format.
• For the ST data, you will need 'ST_data.csv', representing the gene expression matrix of spots, in the 'gene×spot' format.

python geneFilter.py
Rscript makePseudo.R
python getGraph.py
python runModel.py

Then you will get your results in result.csv.