Issues fitting ka parameter

Hello,

I have simbiology 2019a. I am trying to fit ka using a one compartment PK model (I have Vd and Cl fixed).

I had previously scanned my model for ka and found that ka should be approximate 0.5. When I compared simulations to observed data, 0.5 allowed for the model to fit the data well.

I next tried fitting the data using the same concentrations in the model simulations. The ka's were much larger than expected. Upon further expection, the generated simulated concentration vs time profile doesn't produce a curve. It appears as if the model fittings are not generating a concentration profile. I checked if I was making a mistake in the dosing information or response information, but my fitting screen is similar to the example (https://www.mathworks.com/help/simbio/gs/fit-data.html)

Any input or additional references to compare would be greatly appreciated.

Thanks

[Arthur Goldsipe]

Is there any chance you could share your actual SimBiology project with us? That will make it much easier for us to help diagnose the problem. If not, can you give us some more details? Even if you can't share the original data you're using, maybe you could create some "fake data" that illustrates the problem. Otherwise, we'll probably need to ask you a lot of detailed questions to figure out exactly what's going on.

[Brett Fleisher]

Sure, I can provide you with the data. I've attached the file below. My goal is to fit a 1 compartment oral comparment. I've run NCA on the data set and generated an apparent Volume of distribution (661.89 L) and apparent CL (23.39 L/hr). I need to account for bioavailability (F=0.45). Effect compartment and Gut compartments were set to 1 L. The drug concentration I'm estimating is in the central compartment.

My equations to explain this so far are the following:

Fluxes

Absorption = (ka*Drug_Oral_Dose*F)*Gut

Elimination = (CL/Vd*Drug_Central)*Effect

ODEs

d(Drug_Central)/dt = 1/Effect*(Absorption - Elimination)

d(Drug_Oral_Dose)/dt = 1/Gut*(-Absorption)

Thanks in advance for the help

[Peter Bloomingdale]

Just a quick comment that may help you. The way you are modeling bioavailability is essentially just changing the rate of absorption.

Dose*F should be incorporated as the initial condition of the gut compartment.

[Brett Fleisher]

Great, thanks for help.

I had uploaded the wrong data set. I've attached for the corrected version for your reference.

Thanks again!

Answer 1

[Sietse Braakman]

Hi Brett,

There are three ways you can approach this:

1. by assuming that the dose is reduced by a factor F, such as Peter suggested above. However, keep in mind that your absorption rate is dependent on the concentration/amount of the drug_depot species. So by reducing the dose amount by a factor F, you also reduce the initial absorption rate.
2. by assuming that a fraction F of the drug_depot species makes its way to the drug_central species, while a factor (1-F) makes is passed through the intestine without being absorbed. This ensures that the first order rate is not affected by the reduced dose amount. We have an example on bioavailability on our website as well: https://www.mathworks.com/help/simbio/ug/estimating-the-bioavailability-of-a-drug.html.
3. Get both IV and Oral data and approximate F by F = AUC_Oral/AUC_IV, where AUC is the area under the curve of the drug concentration in the central compartment.

Best,

Sietse

Answer 2

[Brett Fleisher]

Hi Sietse,

Thank you again for the help. I've uploaded the model you had helped me with for future reference.

Much appreciated!

Brett