Segger: Reproducible Analyses for Fast and Accurate Cell Segmentation in Imaging-Based Spatial Omics

This repository contains code and analyses for reproducing figures in the preprint Segger: Fast and accurate cell segmentation of imaging-based spatial omics data. Segger is a graph-based segmentation framework that improves transcript assignment in spatial omics data while offering high sensitivity, specificity, and scalability.

Repository Structure

• notebooks_share/ – Jupyter notebooks for reproducing figures and performing key analyses.
• src/ – Python modules for data preprocessing, visualization, and utility functions.

Associated Data

The datasets used in this study are publicly available and can be accessed as follows:

AWS-hosted datasets for this publication

• Xenium Breast Cancer Dataset:
  Download (tar)
• Xenium Colon Dataset:
  Download (tar.gz)
• Xenium NSCLC Dataset:
  Download (tar.gz)

Publicly available Xenium datasets

• Xenium Breast Cancer (10x Genomics):
  Dataset Link
• Xenium Colon (10x Genomics):
  Dataset Link

Download the datasets Run the following commands to download and extract the datasets:

wget https://dp-lab-data-public.s3.us-east-1.amazonaws.com/segger/xenium_breast.tar
wget https://dp-lab-data-public.s3.us-east-1.amazonaws.com/segger/xenium_colon.tar.gz
wget https://dp-lab-data-public.s3.us-east-1.amazonaws.com/segger/xenium_nsclc.tar.gz

tar -xvf xenium_breast.tar
tar -xvzf xenium_colon.tar.gz
tar -xvzf xenium_nsclc.tar.gz

Contact

For questions or issues, please open a GitHub issue or contact the authors of the manuscript.