Spatial Multiomics Analysis – CosMx RNA & Protein

General Information

Generated: 12/12/2024 Last modified: 19/11/2025

This repository contains all scripts, data structures, and figure-generation code used in the study:

“Spatial single-cell multiomics reveals peripheral immune dysfunction in Parkinson’s and inflammatory bowel disease”

The dataset includes measurements acquired using NanoString CosMx™ Spatial Multiomics in 2023, comprising:

• CosMx RNA profiling
• CosMx Protein profiling

Analyses were performed using:

• R ≥ 4.3.2
• Python ≥ 3.8.10

Environment dependencies are provided in:

• requirements_python.txt
• requirements_r.txt

---

File Overview

/data/

Directory reserved for raw input data (not included in this repository).
Users should place their CosMx raw files here to reproduce the analysis.

/analysis/

Contains complete analysis pipelines for both modalities:

• CosMx_Protein/
• CosMx_RNA/

Each includes a detailed README describing preprocessing, QC, annotation, interaction analysis, and intermediate outputs.

/figures/

Code to reproduce every figure included in the manuscript.

Requirements Files

• requirements_python.txt – Python environment (cell–interaction pipeline)
• requirements_r.txt – R environment (full CosMx workflow)

---

Naming Conventions & Repository Structure

This repository uses a structured, modular naming scheme for methodical and reproducible analysis.

Analysis Naming Logic

Scripts follow numeric prefixes to reflect the analysis workflow:

• 0_* → Curation & raw processing
• 1_* → Annotation or normalization (modality-specific)
• 2_* → Cell typing workflows
  • Includes both .R and .py scripts when required
• 3_* → Cell–cell interaction (scotia pipeline)

Additional analysis utilities include:

• abundance_enrichment.R
• scotia_cell_int.py

Data-Type Directories

Subfolders indicate the nature of stored data:

• Files/ – intermediate files
• Molecules/ – molecule-level CosMx outputs
• Polygons/ – segmentation polygons
• Objects/ – Seurat objects or metadata
• Markers/ – marker definitions used in annotation
• Results/ – exported results and final outputs

Figures

Figures follow the convention:

figure1.R, figure2.R, … sup_figN.R

Image exports are stored within:
figures/plots/

---

Full Repository Structure

├── CosMx-Bolen.Rproj
├── LICENSE
├── README.md
├── README.html
├── image.png
├── requirements_python.txt
├── requirements_r.txt
│
├── analysis/
│   ├── CosMx_Protein/
│   │   ├── 0_curation.R
│   │   ├── 1_normalization.R
│   │   ├── 2.0_celltyping.R
│   │   ├── 2.1_celltyping.py
│   │   ├── 2.2_celltyping.R
│   │   ├── 3_cell_interaction.R
│   │   ├── abundances_enrichment.R
│   │   ├── scotia_cell_int.py
│   │   ├── Files/
│   │   ├── Objects/
│   │   │   └── db.csv
│   │   ├── Polygons/
│   │   ├── Results/
│   │   └── Ensembl Cosmx protein.xlsx
│   │
│   └── CosMx_RNA/
│       ├── 0_curation.R
│       ├── 1_annotation_supervised.R
│       ├── 2.normalization.R
│       ├── 3_cell_interaction.R
│       ├── abundance_enrichment.R
│       ├── Files/
│       ├── Markers/
│       ├── Molecules/
│       ├── Objects/
│       ├── Polygons/
│       └── Results/
│
├── data/
│
└── figures/
    ├── figure1.R
    ├── figure2.R
    ├── figure3.R
    ├── figure4.R
    ├── figure6.R
    ├── plots/
    ├── sup_fig1.R
    ├── sup_fig3.R
    ├── sup_fig4.R
    ├── sup_fig5.R
    ├── sup_fig6.R
    └── sup_fig7.R

---

Methodology Overview

Both CosMx RNA and Protein pipelines follow a standardized workflow:

1. Curation

   • Create Seurat objects
   • Extract molecules, polygons, objects into separate folders to maintain structure and reduce memory load

2. Normalization & QC

3. Cell Annotation

   • Supervised annotation workflow

4. Abundance & Enrichment Analysis

5. Cell–Cell Interaction

   • Implemented using the scotia Python-based pipeline (scotia_cell_int.py)

---

Data Access & Funding

This research was funded in whole or in part by Aligning Science Across Parkinson’s (ASAP-020527) through the Michael J. Fox Foundation for Parkinson’s Research (MJFF).

---

License

MIT LICENSE.

Citation

If using data or code from this folder, cite:
Bolen et al., 2025. Spatial single-cell multiomics reveals peripheral immune dysfunction in Parkinson’s and inflammatory bowel disease.