Installation

System Requirements

Python

≥ 3.9, < 3.11

OS

Linux (Ubuntu 22.04 recommended), macOS, or Windows

RAM

16 GB minimum; 32 GB or more recommended for large tissue sections

GPU

A CUDA-capable GPU is required for production CellposeSAM workloads; around 6 GB VRAM is a practical minimum for typical runs. Cellpose2 also benefits from GPU acceleration, and InstanSeg automatically uses CUDA, MPS, or CPU depending on availability

Disk

~5 GB free for model weights, virtual environments, and intermediate outputs

System libraries

libgl, libvips (≥ 8.12)
Docker Image

In order to support a completely reproducible analysis environment, Vizgen provides a Docker image of the Vizgen Post-processing Tool vpt that can be deployed locally, in a high-performance compute (HPC) environment, or in the cloud. The docker image contains vpt and all dependencies, as well as:

  • Nextflow

  • AWS CLI v2

  • Example segmentation algorithm json files

  • Simple Nextflow pipeline suitable for deploying vpt in a HPC environment

The docker image may be downloaded using docker pull:

docker pull vzgdocker/vpt
Python Package Index

The Vizgen Post-processing Tool vpt is available as a Python package. The package may be installed using pip:

pip install vpt

Install the legacy Cellpose and Watershed extras only if you need those specific segmentation families:

pip install vpt[all]

Users encountering difficulty installing the Python package in their local environment are encouraged to try the Docker distribution.

Poetry

The Vizgen Post-processing Tool vpt may be installed from source code by cloning the GitHub Repository and installing using poetry:

git clone https://github.com/Vizgen/vizgen-postprocessing
cd vizgen-postprocessing
poetry install

Install the legacy Cellpose and Watershed extras from a source checkout only if you need those families:

poetry install --all-extras
Segmentation Plugins

pip install vpt[all] (and poetry install --all-extras) install the legacy Cellpose and Watershed segmentation families that ship as optional extras of the vpt package.

Newer segmentation plugins are distributed as separate packages and must be installed individually into the same Python environment as vpt:

pip install vpt-plugin-cellpose2      # Cellpose 2 family
pip install vpt-plugin-cellposesam    # CellposeSAM (Cellpose 4 + SAM) family
pip install vpt-plugin-instanseg      # InstanSeg family

Each plugin self-registers with VPT at install time. No additional configuration is required — once installed, the corresponding segmentation_family value ("Cellpose2", "CellposeSAM", or "InstanSeg") can be used in segmentation specification JSON files.

Repository links and additional plugin documentation:

Each repository README includes additional plugin-specific documentation and an alternative source-based installation path if you prefer to clone the plugin repository rather than install it from PyPI.

Note

CellposeSAM vendors its own copy of the Cellpose v4 inference code, so it can coexist with other Cellpose-based plugins in the same environment without dependency conflicts.

Post-Install Verification

After installation, users can verify that vpt is properly configured by running:

vpt --help

If vpt was installed through poetry or in a virtual environment, the environment may need to be activated. For example:

poetry run vpt --help