napari-omero 0.6.0

napari/OMERO interoperability

napari-omero

This package provides interoperability between the OMERO image management platform, and napari: a fast, multi-dimensional image viewer for python.

It provides a GUI interface for browsing an OMERO instance from within napari, as well as command line interface extensions for both OMERO and napari CLIs.

Features

• GUI interface to browse remote OMERO data, with thumbnail previews.
• Load remote nD images from an OMERO server into napari
  • Planes are loading on demand as sliders are moved ("lazy loading").
  • Loading of pyramidal images as napari multiscale layers
  • OMERO rendering settings (contrast limits, colormaps, active channels, current Z/T position) are applied in napari
• Load ROIs from OMERO server into napari as Shapes or Points
• Upload napari annotation Layers (Labels, Shapes and Points) to OMERO.
• Session management (login memory)

[!NOTE] The user experience when working with remote images, particularly large multiscale (pyramidal) ones, like whole slide images, can be significantly improved by enabling the experimental asynchronous mode (n the GUI in Preferences > Experimental > Render Images Asynchronously or with the environmental variable NAPARI_ASYNC=1).

as a napari dock widget

To launch napari with the OMERO browser added, install this package and run:

napari-omero

The OMERO browser widget can also be manually added to the napari viewer using the Plugins menu or programmatically using:

import napari

viewer = napari.Viewer()
viewer.window.add_plugin_dock_widget('napari-omero')

napari.run()

as a napari reader contribution

This package provides a napari reader contribution that accepts OMERO resources as "proxy strings" (e.g. omero://Image:<ID>) or as OMERO webclient URLS.

import napari
viewer = napari.Viewer()

# omero object identifier string
viewer.open("omero://Image:1", plugin="napari-omero")

# or URLS: https://help.openmicroscopy.org/urls-to-data.html
viewer.open("http://yourdomain.example.org/omero/webclient/?show=image-314", plugin="napari-omero")

these will also work on the napari command line interface, e.g.:

# quotes are needed if using zsh
napari "omero://Image:1"
# or
napari "http://yourdomain.example.org/omero/webclient/?show=image-314"

as an OMERO CLI plugin

This package also serves as a plugin to the OMERO CLI

omero napari view Image:1

• ROIs created in napari can be saved back to OMERO via a "Save ROIs" button.
• napari viewer console has BlitzGateway 'conn' and 'omero_image' in context.

installation

While this package supports anything above python 3.9, in practice, python support is limited by omero-py and zeroc-ice, compatibility, which is limited to python <=3.12 at the time of writing.

from conda

It's easiest to install omero-py from conda, so the recommended procedure is to install everything from conda, using the conda-forge channel. For example, to install the plugin, napari, and the default Qt backend, use:

conda install -c conda-forge napari-omero pyqt

from pip

napari-omero itself can be installed from pip, but you will still need omero-py

conda create -n omero -c conda-forge python=3.10 omero-py
conda activate omero
pip install napari-omero[all]  # the [all] here is the same as `napari[all]`

issues

❗	This is alpha software & some things will be broken or sub-optimal!

• experimental & definitely still buggy! Bug reports are welcome!
• remote loading can be very slow still... though this is not strictly an issue of this plugin. Datasets are wrapped as delayed dask stacks, and remote data fetching time can be significant. Enabling asynchronous rendering in napari improves the subjective performance... but remote data loading will likely always be a limitation here. To try asyncronous loading, start the program with NAPARI_ASYNC=1 napari-omero or look in the Preferences on the Experimental tab. Also, keep an eye on the napari progressive loading implementation progress.
• For plugin developers: As napari-OMERO provides images as lazily-loaded dask arrays, napari-plugins need to account for this when retrieving data from napari layers. Keep in mind that forwarding the data to processing steps in plugins may lead to signficant loading and processing times.

contributing

Contributions are welcome! To get setup with a development environment:

# clone this repo:
git clone https://github.com/tlambert03/napari-omero.git
# change into the new directory
cd napari-omero
# create conda environment
conda env create -n napari-omero python=3.10 omero-py
# activate the new env
conda activate napari-omero

# install in editable mode with dev dependencies
pip install -e ".[dev]"      # quotes are needed on zsh

To maintain good code quality, this repo uses ruff, mypy.

To enforce code quality when you commit code, you can install pre-commit

# install pre-commit which will run code checks prior to commits
pre-commit install

The original OMERO data loader and CLI extension was created by Will Moore.

The napari reader plugin and GUI browser was created by Talley Lambert

release

To push a release to PyPI, one of the maintainers needs to do, for example:

git tag -a v0.2.0 -m v0.2.0
git push upstream --follow-tags

Then, the workflow should handle everything!