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

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

For detailed documentation, see README.pdf