Tips for my scientific visualization use case?

Currently, I use Python+Cairo to generate pictures of groups of polygons which represent biological cells:

I am interested in doing something similar using Blender, for two reasons:

  1. The Cairo-based pipeline feels slow.

  2. I am going to be implementing 3D cells (3D polyhedra), and it will not be so easy to use Cairo to visualize these.

Based on the tinkering I have done with Blender so far, I plan on:

  1. Using Blender’s integrated Python to import time series (using the srsly module to deserialize input data) data for the cells.

  2. Scripting meshes representing the cells based on the imported data.

  3. Scripting lines to represent the “sticks”, etc. Circles for “target zones”, and so on.

  4. Using the Freestyle Renderer to create the final images.

What do you think of this plan overall? Are there some Blender features I should look into in particular which might be helpful for my project?

Dupliverts for putting the sticks on the surface of a geo sphere. Or other geometry that ll represent your cells.

There might be already solutions out there, did comes with a quick search: https://cellmicrocosmos.org/index.php/cm5-project

Instead of freestyle rendering, I found something much better! GreasePencil!

In particular, how to script using it: https://towardsdatascience.com/blender-2-8-grease-pencil-scripting-and-generative-art-cbbfd3967590