Biocellion workflows

[Robert Yaman]

Hi all,

I've just begun tinkering with Biocellion, and I'm wondering if anybody has any tips on how to make workflows more convenient. Currently, I don't have access to a linux machine, so I'm running Biocellion on a VM in GCP. Developing directly on the VM isn't great since there is some network lag, and I don't have access to my IDEs. It's also a bit inconvenient to develop locally since I have to move files to and from my VM in order to see if the code compiles / does what I want. I'm wondering if anyone has any tips/tricks for speeding up development on a mac. 

Additionally, I'm wondering how people generally build complex models. Do you generally start with smaller parts of the model, compile and test, then incrementally build out the complexity? How useful have you found writing tests in order to be sure that the model is doing what you think it's doing?

Any advice is greatly appreciated!

Robert

[seunghwa.kang]

Hello Robert,

Welcome to the Biocellion community!!!

I don't have much to say about the first question. At this moment, I don't have a good answer besides finding a Linux box or you may reconfigure your mac to support dual boot
.

Regarding the second question, complex models typically have multiple submodels which can also be complex. And I typically start with a combination of multiple very simple submodels and update individual submodels incrementally. In many cases, there are  a lot in common in submodels of different biological systems, and if this is the case, I adopt an existing submodel and tune this submodel for the new biological system (for example, modeling biomechanics of skin is not completely different from modeling biomechanics of gut or bacterial biofilm). I often find ways to improve the adopted submodel in tuning and validation processes, and this often benefits multiple biological system models.

And I think it is very important to fully test individual submodels if you wish to build a complex model. Debugging becomes more and more difficult as your models become more and more complex. If you can trust your submodels, you just need to validate interactions among submodels, but if you don't you need to suspect everything and if you have enough experience in software engineering, you must know how difficult it is to integrate crappy software modules; pretty similar for biological system modeling.

One of the Biocellion's main design goal is to facilitate this process. Biocellion API is designed to accommodate mathematical models of multiple different biological systems to facilitate submodel reuses and I am continuously adding mathematical libraries to Biocellion so all the Biocellion users can collaboratively test and validate those libraries.

Thank you very much,

-seunghwa

[seunghwa.kang]

And you may find the following link useful.

https://www.rosaandco.com/webinars/2017/development-of-a-multiscale-skin-barrier-model-for-de-novo-in-silico-prediction

The above describes a skin model developed on Biocellion, and it might be the most complex model developed on Biocellion (excluding proprietary models under development).

-seunghwa

On Tuesday, January 23, 2018 at 3:59:33 PM UTC-8, Robert Yaman wrote:

[Robert Yaman]

Thanks a lot for the pointers!