Cellpose is a general purpose, deep learning-based segmentation algorithm for biological images (Cellpose on GitHub).
This folder contains guides and resources to help you use Cellpose on the Crick’s HPC cluster, with and without Jupyter notebooks, plus demo data and example notebooks.
Getting Started
- Choose your installation method:
- Want to use Cellpose in a Jupyter notebook on the cluster? 👉 Follow these instructions.
- Want to run Cellpose from FIJI or as a GUI in an OnDemand Desktop session (without Jupyter notebooks)? 👉 See these instructions.
⚠️ Be sure to use a GPU node while installing on the HPC cluster (nemo)!
- Try the Example Notebook:
- Cellpose GPU Check Notebook:
Quickly verify your Cellpose installation and GPU access. - Crick Cellpose-SAM Demo Notebook:
Step-by-step guide to segmenting images with Cellpose-SAM on Nemo. Works with 2D images.
- Cellpose GPU Check Notebook:
- Download Demo Images:
Try out Cellpose using our demo images.
File & Folder Overview
-
demo_images/
Example images to try with Cellpose. -
Cellpose_GPU_Check.ipynb
Jupyter notebook for testing your Cellpose install and GPU setup. -
Crick_CellPose_SAM.ipynb
Full workflow notebook for segmenting images with Cellpose-SAM. -
install_cellpose_nemo_jupyter.md
Step-by-step guide to installing Cellpose in a JupyterLab environment on Nemo. -
Run_Cellpose_From_FIJI_on_OnDemand.md
Guide to installing Cellpose for use from an OnDemand desktop session, including FIJI integration. -
install_cellpose_0525.sh
Shell script for setting up the Cellpose environment.
❓ Need Help?
Issues or questions about the install? 3D images? Advanced workflows?:
The provided notebooks are for 2D images. For 3D support, or further help, contact Todd, Dave, or Sara from CALM.
Last updated: May 2025