Development of a pipeline for spatial transcriptome analysis of samples from Hodgkin's lymphoma patients

Abstract

OVERVIEW Automated pipeline for spatial transcriptomics analysis of classical Hodgkin lymphoma samples using 10x Genomics Visium low-density platform. Includes preprocessing (Space Ranger), quality control, batch correction (Harmony, scVI), and downstream analysis.

Master's thesis, Justus Liebig University Giessen (2026).

CONTENTS - Bash script for automated Space Ranger preprocessing - JupyterLab notebooks for quality control, batch correction, and downstream analysis - PDF reports documenting all analysis steps and results - HTML tables with QC metrics, correlation matrices, and differential expression results - Figures and visualizations from all analyses - Test dataset for pipeline validation (mouse spleen, GEO: GSE254652)

TECHNICAL REQUIREMENTS - Ubuntu 24.04.2 LTS, Space Ranger v3.1.3, Python 3.12.10 - Python packages: Scanpy (v1.10.4), Squidpy (v1.6.5), harmonypy (v0.0.10), scVI (v1.3.1.post1) - Hardware: Minimum 40 GB RAM, 4 CPU cores recommended - Reference files: refdata-gex-GRCh38-2020-A, Visium_Human_Transcriptome_Probe_Set_v2.0_GRCh38-2020-A

USAGE 1. Prepare input files (SampleSheet.csv, Aggregation.csv) 2. Configure paths in script.sh and collect_output.py 3. Run preprocessing: bash script.sh 4. Generate metadata: python collect_output.py 5. Update base paths in JupyterLab notebooks 6. Execute JupyterLab notebooks sequentially

For detailed documentation, see README.pdf