Clear Filters
Clear Filters

Data fit to observables

5 views (last 30 days)
Justine on 8 Apr 2024
Commented: Arthur Goldsipe on 10 Apr 2024
I have built a PBPK model on Simbiology (2023b), with several compartments, representing individual organs. Each compartment is then subdivided into vascular, interstitial, intracellular. When simulating the model, Simbiology computes then the concentrations over time per subcompartment. As I have data on the organ level, I would like to fit my data to the model (several organ at a time). I have then "pooled" my concentrations per organ. For instance, for liver: liver_vascular, liver_intracellular and liver_interstitial, that I have pooled into liver_all as an observable to represent the concentrations in the whole organ. For simulation step, I don't encounter any issue.
To fit the model to my data, I tried the "fit data" programm from the model analyzer application, but when it comes to data mapping, I don't see the option to choose my observables in the drop down menu (although I added them to the states to log). I am mostly using the application, and not so much coding here. Any tips and tricks on how to fit my data to observables will be highly appreciated. Unfortunately I can't share my project file.
Thank you very much in advance!

Accepted Answer

Arthur Goldsipe
Arthur Goldsipe on 8 Apr 2024
Hi Justine,
I just tried this in R2023b, and I was able to select an observable in the mapping for the response. I'm attaching a sample project that does that.
If that still doesn't help you resolve your problem, then it may be a specific to your project. Since you cannot share your project publicly, it will be harder to help you. Here are some of the options we've used when confidentiality is an issue:
  • Share the project privately with MathWorks, with the understanding that we will delete it when we're done debugging.
  • Similarly, some pharaceutical companies have set up an NDA with MathWorks so that we can safely share models. If you work for a pharmaecutical company, it may be worth seeing if we already have such an NDA in place.
  • Some users are able to set up a video call where we interactively try to debug the problem. MathWorks will still "see" the model, but we won't actually have the any files.
  • Remove any confidential information from the model before sharing it. The SimBiology team has a tool we can share to help with that. (It renames all the components in the model. The structure of the model is still visible, but usually obfuscating names is sufficient for addressing IP concerns.)
  • Create standalone reproduction steps for the problem that don't involve any of the confidential information.
Please feel free to contact me via my MATLAB profile if you want to explore any of the options that involve direct communcation with MathWorks.
Justine on 10 Apr 2024
Thanks a lot for the quick answer Arthur!
By looking at your file, the name of the observable still doesn't appear in the drop down menu but by just writing it and running the program, it works fine. Thanks again!
Best regards
Arthur Goldsipe
Arthur Goldsipe on 10 Apr 2024
Oh, good point. I saw the observable in the dropdown after I typed it in manually. But I didn't realize it wasn't in the list before I typed it in manually. I'll add that to our bug/enhancement database so that we can get it fixed in a future release.

Sign in to comment.

More Answers (0)


More Answers in the  SimBiology Community


Find more on Scan Parameter Ranges in Help Center and File Exchange





Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!