MultiModal Big Image Data Sharing and Exploration

MoBIE is a framework for sharing and exploring large multi-modal image datasets. Example projects:

• A cellular atlas for Platynereis dumerilii, the PlatyBrowser
• Morphogical changes in SARS-CoV2 infected cells in EM Tomograms and Volume EM.

• NEUBIAS Academy MoBIE Webinar
• Object browsing and attribute colouring
• Exploring gene expression in Platynereis atlas
• Visualising gene expression and chromatin segmentation in Platynereis atlas

For now, we provide two software tools to explore and generate MoBIE projects:

• mobie-viewer-fiji is a BigDataViewer based Fiji plugin to explore MoBIE projects from local or remote sources.
• mobie-utils-python is a python library to generate data in the MoBIE data storage layout.

Tutorials for using MoBIE are available at http://mobie.github.io/, and installation instructions here

To be accessible via MoBIE tools, your data needs to be organized as a project. One project can contain several datasets.

See the folder /data in this repository for an example project. It uses a small subset of the data from the PlatyBrowser.

A project consists of a root folder with the file datasets.json that lists the available datasets and the datasets, each stored in a separate subfolder.

+datasets.json
+dataset1/
+dataset2/

A dataset folder is structured as follows:

+images/
|  +--images.json
|  +--local/
|  +--remote/
+misc/
|  +--bookmarks
+tables/
+README.txt

with the following elements:

• images/images.json lists the avaialable image data and stores their display options
• images/local contains the metadata for image data stored locally, in bdv.xml data format
• images/remote contains the metadata for image data stored remotely (s3 compatible cloud store) in bdv.xml data format
• misc contains miscellaneous data
• misc/bookmarks contains the bookmarks, stored as json files
• tables contains tabular data associated with image segmentation data
• README.txt gives a description of this dataset (optional)

MoBIE uses the BigDataViewer file format to represent image data, either stored locally or on a s3 object store:

• bdv.n5 for local data
• bdv.n5.s3 for data stored on a s3 compatible object store. Note that this is not part of the official bdv spec yet, we are working towards merging it

The file images.json lists all image data that is available for a dataset. Currently, we support three different types of image data:

• image: greyscale image data
• segmentation: segmentaiton image data, where each segment is assigned a unique id; supports tables
• mask: binary mask

The bookmark folder must contain a file default.json that gives the bookmark for the initial view when opening the dataset. It may contain additional bookmarks stored as json.

Tables are stored as tab separated values and can be read from the filesystem or git. For each segmentation with associated tables, the tables are stored in tables/<IMAGE-NAME>/. This folder must contain a table called default.csv, it can contain additional tables. All tables must contain the column label_id linking its rows to objects in the segmentation.

MoBIE is primarily deveopled by Christian Tischer (@tischi), Kimberly Meechan (@K-Meech) and Constantin Pape (@constantinpape). It has been initially developed for the PlatyBrowser but has since grown in scope. In addition, Hernando Martinez Vergara (@HernandoMV), Martin Schorb (@martinschorb) and Valentyna Zinchenko (@vzinche) have contributed to the development.

If you use MoBIE for your research, please cite Whole-body integration of gene expression and single-cell morphology.