SpaceNet: A pathology-driven spatial disease network framework for exploring the molecular changes of lesions

SpaceNet, a pathology-guided spatial network that combines imaging data with single-cell/single-nucleus RNA sequencing (sc/snRNA-seq) and spatially resolved transcriptomics (SRT). SpaceNet identifies regions of interest (ROIs) directly from pathology or imaging features and examines these areas to find enriched cell types, abnormal gene activity, and spatial protein-protein interaction (PPI) networks.

This repository describes how to analyse your data with SpaceNet.

Details on SpaceNet:

• Installation
• Usage
• Relavants

---

Quick start

1. Installation of SpaceNet

#installation of conda_env, recommended version python3.8.20, consistent with development env
cd SpaceNet-release
conda env create -n spacenet python=3.8.20 
conda activate spacenet

#installation of requirements
pip install -r requirements.txt

2. Accessibility to other softwares/database

choose one to integrate your scRNA-seq data with SRT data

• CellTrek
• Cell2location
• Tangram

visit STRING to obtain the lastest complete PPI database, or use a quick-start version of PPI database (support species: human, mouse, rat)

• STRING (Optional)

visualize your result with cytoscape

• Cytoscape

3. Define ROI in your SRT data

use our tool to select ROI, or save your disease_metric in st_adata.obs

4. Running the SpaceNet pipeline on the [example dataset]

example datasets can be accessed here

• dot_like_example
• band_like_example

5. Expected output

The output of SpaceNet containing:

• file 1: A csv file of enriched cell-types and theirs enrichment score
• file 2: A csv file of enriched genes and theirs enrichment score
• file 3: A csv file of spatial protein-protein interactions net
• file 4: A csv file of gene-celltype net

---

Relavants and more information

Should you have any questions, please contact Menglei Wang at wangml@zju.edu.cn or Hudong Bao at ProDong@zju.edu.cn