Create google colab 'sandbox' environment for Biosimulators interaction

[AlexPatrie]

Problem scenario

• A user with a non(or sparsely)-programmatic background, such as a student, wants to use our tooling. They are not familiar with Docker, nor are they familiar with Python on the level of CLI interaction. Their professor speaks of biological simulations, which piques the interest of the student. They want to learn the concepts mentioned in class while becoming familiar with programming.

---

Addresses a need for the following:

• A 'startup' environment that users are able share and access with a web browser that is lightweight

---

What does this PR do?

• Adjusts Dockerfile for wheel build dependency issue
• Adds a jupyter notebook entitled "Biosimulators_Playground" which will be formatted into a Google Colab notebook that is viewable on this repo's landing page

---

To Do PRIOR to merge

• Format for Google Colab
• Refactor for build error

[netlify]

❌ Deploy Preview for determined-booth-118c06 failed.

Name	Link
🔨 Latest commit	5583e2d
🔍 Latest deploy log	https://app.netlify.com/sites/determined-booth-118c06/deploys/644fcf339a9819000833f91f

[AlexPatrie]

During CI, an old version of utils is being used to build the tutorials repo and thus there is an error when importing the object CompilationException. I am going to attempt to assert higher upper bounds for the biosimulators-utils dependency requirement. After all, this is the tutorial and should have an up-to-date utils backend!

[AlexPatrie]

I propose a change to the readme as follows:

BioSimulators tutorials

This repository contains Jupyter notebooks with tutorials for the Python APIs for the simulation tools registered with BioSimulators and a Jupyter notebook server for running the notebooks. NEW: As an alternative to the Jupyter Notebooks and relative Jupyter Notebook server, we have added a Google Colab Notebook for users that are more comfortable with a shareable, annotatable "sandbox environment". Complete with the required dependencies, input data, and explicitly-defined methods that are essential to interacting with Biosimulators, this Google Colab notebook serves as "click and play" experience.

The tutorials are located in the tutorials subdirectory. The README in this directory summarizes the content of the tutorials.

Getting started

Running the tutorials online with Google Colab (web-based, minimal setup)

Open https://colab.research.google.com/drive/1aI7Z6ZBcQzOc9PJllW9N9oeTcFtXfhTd#offline=true&sandboxMode=true in your browser or click on the above Colab badge.

Running the tutorials online with Binder

The easiest way to run the tutorials is with Binder. Simply open https://tutorial.biosimulators.org/ in your browser.

Running the tutorials on your own machine

Run the following command to download the tutorials:

docker pull ghcr.io/biosimulators/tutorials

Run the following command to launch a Jupyter notebook server at http://localhost:8888 for the tutorials in this repository:

docker run -it --rm -p 8888:8888 ghcr.io/biosimulators/tutorials jupyter notebook --ip=0.0.0.0 --port=8888

Alternatively, run the following command to launch a Jupyter lab server at http://localhost:8888 for the tutorials in this repository:

docker run -it --rm -p 8888:8888 ghcr.io/biosimulators/tutorials jupyter lab --ip=0.0.0.0 --port=8888

License

This package is released under the MIT license.

Development team

This package was developed by the Karr Lab at the Icahn School of Medicine at Mount Sinai in New York as part of the Center for Reproducible Biomodeling Modeling with assistance from the contributors listed here.

• Bilal Shaikh
• Jonathan Karr

Funding

This package was developed with support from the National Institute for Bioimaging and Bioengineering (award P41EB023912).

Questions and comments

Please contact us at info@biosimulators.org with any questions or comments.