A plugin to visualize 3D single cell omics
A plugin to visualise 3D spatial single cell omics
Test and atlas datasets¶
Because the datasets representing the mouse embryo at stages E8.5 and E9.0 are rather large, it is not possible to host them on GitHub. They are instead hosted on figshare at the following links:
Once downloaded, one can open them in the viewer as explained below (note that the files for the tissue names are stored in the json file there:
napari-sc3D-viewer/test_data/corresptissues.json). It can be downloaded by right-clicking on the following link and then clicking on "Save link as".
Disclaimer: While we tried to make the installation and usage as easy as possible, please keep in mind that napari-sc3d-viewer is still under development, it has been and is being developed by a single person. We will be happy to answer any question and help in any way.
There are many ways to install our viewer, but the global idea is that it works in two steps:
If you have decided to use command line, as napari developers do, we strongly recommend to install the viewer in an environement such as a conda environment
conda for example:
conda create -n sc3D python=3.10
conda activate sc3D
Quick trouble shooting¶
conda create -n sc3D python=3.10
conda activate sc3D
conda install pyqt pip
pip install napari
Installation via graphical interface¶
Installation via the terminal¶
conda install napari-sc3d-viewer
pip install napari-sc3d-viewer
Finally, to install latest development version :
pip install git+https://github.com/GuignardLab/napari-sc3D-viewer.git
Installation of the surface computation module¶
To install the surface computation enabled version it is necessary to use Python 3.9 (until VTK is ported to Python 3.10) and you can run one of the following commands:
pip install '.[pyvista]'
from the correct folder or
pip install 'napari-sc3D-viewer[pyvista]'
conda install 'napari-sc3D-viewer[pyvista]'
to install directly from pip or
pip install 'napari-sc3D-viewer[pyvista] @ git+https://github.com/GuignardLab/napari-sc3D-viewer.git'
to install the latest version
napari-sc3D-viewer allows users to easily visualise and navigate 3D spatial single-cell transcriptomics using napari.
Starting the plugin¶
First, you need to start napari, for example, one can start it from a terminal just by typing:
in the correct environment.
Then, one can follow the following steps to browse the dataset.
To open the plugin you can click on the "Load spatial single cell" from the
Plugins -> napari-sc3d-viewer menu:
Once opened you should have an interface poping similar to the one showed in the image below (note that it might not be exactly the same depending on the version of the viewer you are using).
Loading and opening a dataset¶
The json file should look like that:
"10": "Anterior neuroectoderm"
If no json file or a wrong json file is given, the original cluster id numbers are used.
The h5ad file should be informed in (1) and the json file in (2).
data be your h5ad data structure. To work properly, the viewer is expecting 4 different columns to be present in the h5ad file:
- the cluster id column (by default named 'predicted.id' that can be accessed as
- the 3D position column (by default named 'X_spatial_registered' that can be accessed as
- the gene names if not already in the column name (by default named 'feature_name' that can be accessed as
- umap coordinates (by default named 'X_umap' that can be accessed as
If the default column names are not consistent with your dataset, they can be changed in the tab
Parameters (3) next to the tab
Once all the data paths and fields are correctly informed pressing the
Load Atlas button (4) will load the dataset.
Exploring a dataset¶
Once the dataset is loaded there are few options to explore it.
The viewer should look like to the following:
It is divided in two main parts, the Tissue visualisation (1) part and the Metric visualisation (2) one. Both of them are themselves split in two and three tabs respectively. All these tabs allow you to visualise and explore the dataset in different fashions.
The Tissues tab (1.1) allows to select the tissues to display, to show the legend and to colour the cells according to their tissue types.
The Surfaces tab (1.2) allows to construct coarse surfaces of tissues and to display them.
The Single metric tab (2.1) allows to display a metric, whether it is a gene intensity or a numerical metric that is embedded in the visualised dataset. This tab also allows to threshold cells according to the viewed metric, to change the contrast and the colour map.
The 2 Genes (2.2) tab allows to display gene coexpression.
The umap tab (2.3) allows to display the umap of the selected cells and to manually select subcategories of cells to be displayed.
The plugin is meant to be easy to use. That means that you should be able to play with it and figure things out by yourself.
That being said, it is not always that easy. You can find below a series of videos showing how to perform some of the main features.
Displaying one gene¶
Displaying two genes co-expression¶
Playing with the umap¶
Computing and processing the surface¶
Contributions are very welcome. Tests can be run with tox, please ensure the coverage at least stays the same before you submit a pull request.
Distributed under the terms of the MIT license, "napari-sc3D-viewer" is free and open source software
If you encounter any problems, please file an issue along with a detailed description.
- 24 July 2023
- 21 June 2022
- Information not submitted
- Stars: 18
- Forks: 1
- Issues + PRs: 0