Update script-to-generate-datasets.sh to use genSamples.py

[adyprat]

No description provided.

[amoghpj]

Fixed in 992d701.

Hi @matthieubulte , we realized that the previous scripts-to-generate-datasets.sh was missing a way to generate samples of cells, and that script had not been included in the repository. This has now been fixed (See scripts/genSamples.py). I have updated the shell script with these changes, and it now does the following:

1. Do 600 simulations. All simulated trajectories are stored in the /simulations folders
2. Generate one sample of 500 cells
   In the case of the Curated model, additionally:
3. Run SlingShot on this sample of 500 cells.
4. Additionally generate dropouts.

I would greatly appreciate it if you tested this script and gave me some feedback in terms of functionality and/or documentation.

We appreciate your efforts to make BoolODE more usable. Thank you so much for your pull requests!

cc @adyprat

[matthieubulte]

Hi, that's a nice update! Since you ask me, here are my comments on what could be improved:

• There is still the error on Mac OS because of the plotting by default. I'm sure that this will be fixed, but it would be sad to forget and make the script unusable for Mac users!

• I'm not sure if you should remove the docker instance (or provide a docker-less option) or go full docker to make sure the user doesn't even have to install any of your Python dependencies. I think this is something that you could look into if you want to improve the user experience.

• The README at the moment is great, but it would be even better if you added some documentation about the script-to-generate-datasets.sh, or at least mention that it can be used to reproduce the datasets you are using in your Beeline paper.

Anyways, thanks for this update!