Simultaneous Fit of IV and PO data using SimBiology

Hello,

I was wondering if there was a way to model both IV and PO data simultaneously for the same patients using SimBiology; I've successfully fitted the IV data in one model and the PO data using the same model but with different initial estimates, but I was wondering if it's possible to do one fit for both datasets. I get an error when I try to do both because the time course for the IV data and the PO data uses the same scale (0-144 hours)

Thanks in advance!

Answer 1

[Arthur Goldsipe]

Yes, it is possible to fit two model outputs at the same time. Do you want to do this using sbiofit or a fit program in the SimBiology Analysis app? Here's an example of a simultaneous fit to two different compartments.

I don't understand what you mean about getting an error because the data uses the same time scale. If you can share some more details (the exact error message and how to reproduce it), then we should be able to help you.

[Engie Salama]

Hi Arthur,

Thanks so much for responding. I was originally using the fit program in the SimBiology Analysis, and I've attached a screenshot of the error message. The dataset structure is that for each patient, I have the IV dose then the IV time points, then the PO dose and the PO time points (0-144 for both). Any help would be appreciated!

[Arthur Goldsipe]

Ah, this error means that you need to sort your data so that time is increasing. You will need to do that step outside of the SimBiology Analysis app before importing your data.

[Engie Salama]

I see, I was wondering if it was possible to do something similar to this bioavailability example but with multiple patients:

https://www.mathworks.com/help/simbio/ug/estimating-the-bioavailability-of-a-drug.html

In that example, the time course of 0 to 16 hours is the same for both the IV and PO groups; I want to do this but with multiple individuals.

[Arthur Goldsipe]

Yes, that is possible. You just need to make sure that for each patient, earlier times appear in earlier rows in the data. To do this, I usually sort by patient and then by time for each patient. But you could also just sort the entire data set by time, if you prefer.

[Engie Salama]

Thank you! It was able to run when I sorted the data as you suggested, but then I got one fit for each patient even though there are two curves: one for IV and one for PO. Is there a way to get fits for each curve separately? At this point, I was just going to rename the patient IDs as 100001A for IV and 100001B for PO...

[Arthur Goldsipe]

Ah, sorry. I thought IV and PO represented different measurements taken during the same treatment of a patient. I looked up the term "PO" and now understand that this is a different dosing administration scheme and is associate with a different time course of measurements.

In that case, I would do exactly as you propose: Create a new label that is unique for each combination of patient and dose.

[Engie Salama]

Excellent, thank you for your help Arthur :)