2024-jn-omero-pipeline

Material and solutions for the course "Bioimage data management and analysis with OMERO" held in Heidelberg (13th May 2024)

Module 3 (1.45 pm - 3.45 pm): OMERO and Jupyter Notebooks

Instructor - Dr Riccardo Massei

Jupyter Notebook Pipeline Overview

Main goal of the workflow is to show the potential of JN to perform reproducible image analysis in connection with an OMERO instance. In this specific example, we are performing a simple nuclei segmentation from raw images uploaded in OMERO.

Overview on the JN pipeline:

1. Connect to OMERO using the python API
2. Fetch an image for image analysis
3. Perform the following pre and post-processing steps

3a) Denoising 3b) Binarization 3c) Labelling 3d) Feature extraction 4) Push Results back to OMERO by adding metadata and additional informations

Furthermore, it will be show how to process can be automatized by the execution of automatic loops throught the OMERO datase. Most of the pre and post-processing step were created by taking inspiration from the Bio-image Analysis Notebooks. We thanks Robert Haase for creating this material.

Course Preparation and checklist

---

Connect to the OMERO instance

You can 1) connect to your institutional instance, 2) run OMERO.server and OMERO.web using docker-compose locally or 3) use temporary OMERO account prepare by Müster University which can be accessed and used the day of the course. More information regarding these accounts will be give the day of the course.

IMPORTANT - Installing omero-py

Please, start to create your environment by installing omero-py in python 3.8 using the command:

conda create -n myenv -c conda-forge python=3.8 omero-py

you can then install the other packages into this environment

Run the Jupyter Notebook

Jupyter is open-source software and service for interactive computing.

We suggest to install the conda environment before the course and run JN locally on your personal computer.

Optionally, you can also use the service GoogleColab. Please, install your conda environment with omero-py writing the following command:

 pip install https://github.com/glencoesoftware/zeroc-ice-py-linux-x86_64/releases/download/20240202/zeroc_ice-3.6.5-cp310-cp310-manylinux_2_28_x86_64.whl 
 pip install omero-py==5.19.1 

Install then all the other packages.

Setup the conda environment

• You can set-up your environment using the py_pipeline.yml or env.yml before the course. In case of problems, please contact the organizators.

Examplary dataset for the present course...

We used image set BBBC014v1 provided by Ilya Ravkin, available from the Broad Bioimage Benchmark Collection [Ljosa et al., Nature Methods, 2012]. For details and biological background see https://www.broadinstitute.org/bbbc/BBBC014/. Images can be founded in this repository at 01_data/HCS_data.zip. Data need to uploaded into your OMERO instance before the exercise. Otherwise, data will be already uploaded in the prepared server

..or bring your own data :)

In case you have your own data and want to perform a basic nuclei segmentation, your dataset is welcome. In case, you can also browse IDR for potential and interesting case studies.

Useful resources before the course:

• Jupyter Notebook Webpage:

https://jupyter.org/

• OMERO Python API:

https://docs.openmicroscopy.org/omero/5.6.0/developers/Python.html

• Bio-image Analysis Notebooks:

https://haesleinhuepf.github.io/BioImageAnalysisNotebooks/intro.html

https://scikit-image.org/