cSCC continuum analyses

This repository contains the code required to reproduce the end-to-end analyses of HTG EdgeSeq (bulk RNA profiling) and GeoMx DSP (spatial transcriptomics) data from the publication:

Oterino-Sogo, S. & Naji, F. et al. Spatial and bulk transcriptomic profiling defines the molecular evolution of cutaneous squamous cell carcinoma and reveals stage-specific biomarkers of clinical relevance. bioRxiv (2026). https://doi.org/10.64898/2026.04.30.721943

Data Availability

The analyses rely on processed expression matrices derived from the following GEO datasets:

• HTG EdgeSeq (HTG-seq): GSE319969
• GeoMx DSP: GSE319968

⚠️ The scripts in this repository are designed to work with processed data files, not raw FASTQ/SRA files. Raw sequencing files are available through the SRA links provided within each GEO entry.

Repository Structure

The repository contains two main directories:

• scripts/
  • Differential expression, clustering, and visualization analyses
• gprofiler2_ORSUM/
  • Pathway enrichment and ORSUM summarization workflows

Reproducing Differential Expression and Clustering Analyses

This section describes how to reproduce:

• Differential expression analyses (LRT and Wald tests by stage; DESeq2-based)
• LRT clustering analysis using DEGreport

Required Input Files

1. We provide a link to download the input files from our local storage to agilize reproducibility. You may also download the processed datasets from GEO (see Data Availability).
2. Place all required input files in the directory specified by path_to_input.
3. Download the file: Heatmaps_gene_list.xlsx and place it in the same input directory.

⚠️ The scripts will not run without these files.

1. Clone the Repository

git clone https://github.com/cbib/cSCC_continuum_analyses
cd cSCC_continuum_analyses

2. Download processed files

This folder contains the processed count matrices and Heatmaps_gene_list.xlsx file.

wget --no-check-certificate -r -np -nH --cut-dirs=1 -R "index.html*" http://services.cbib.u-bordeaux.fr/cSCC_gene_tables/data/

3. Create the Conda Environment

conda create --file geomx.yml
conda activate geomx_env

4. Configure Input and Output Paths

Edit the following scripts:

• GeoMx_lrt_reproduce.R
• HTGseq_lrt_reproduce.R

Update the variables:

• path_to_input
• path_to_output

5. Run the Analyses

Rscript GeoMx_lrt_reproduce.R
Rscript HTGseq_lrt_reproduce.R

Output Structure

All results will be written to the directory specified in path_to_output with the following structure:

• DE_genes_by_stage/
  • DESeq2 pairwise comparisons for HTG-seq, GeoMx Macrophages, GeoMx PanCK.
• Heatmap_plots/
  • Heatmaps summarizing pathway enrichment results obtained using manual curation of gprofiler2 and ORSUM output.
• Single_gene_plots/
  • Strip plots for candidate stage-specific cSCC biomarkers identified in the study.

Additionally, the following files are saved directly in path_to_output:

• .png Likelihood Ratio Test (LRT) plots
• .csv files containing gene groups

Reproducing gprofiler2 and ORSUM Pathway Analysis

If you wish to reproduce gprofiler2 pathway enrichment and ORSUM summarization, follow the steps below. ⚠️ These scripts are configured to run on a SLURM-based HPC cluster.

1. Create the Environment

conda create --file gprofiler2_orsum.yml
conda activate gprof_orsum_env

2. Input Requirements

The script gprofiler2_ORSUM/gprof_orsum.sh expects a .tsv file as input and calls gprofiler2_ORSUM/gprofiler_standalone.R

Notes:

• LRT and DEGreport outputs are saved as .csv. They need to be converted to .tsv
• For GeoMx PanCK results, only a subset of gene groups was used for pathway analysis.
• ORSUM requires .gmt files, which can be downloaded here.
• Before running gprof_orsum.sh rename file prefixes by replacing : with _ (example: GO:BP → GO_BP)

3. Run on SLURM

cd gprofiler2_ORSUM/

sbatch gprof_orsum.sh -i Publication_HTG_Edgeseq_gene_groups.tsv -o /results -gmt /gprofiler/gmt/hsapiens -os '500' -org 'hsapiens'

sbatch gprof_orsum.sh -i Publication_PanCK_gene_groups.tsv -o /results -gmt /gprofiler/gmt/hsapiens -os '500' -org 'hsapiens'

sbatch gprof_orsum.sh -i Publication_Macrophages_gene_groups.tsv -o /results -gmt /gprofiler/gmt/hsapiens -os '500' -org 'hsapiens'

Parameters:

• -i : input gene group file
• -o : output directory
• -gmt : path to GMT files
• -os : ORSUM size threshold
• -org : organism (e.g., hsapiens)

Contact and Support

The code in this repository was developed by Sergio Oterino Sogo.

LinkedIn: Sergio Oterino Sogo, PhD.

For reproducibility issues, please open a GitHub issue.