2 replies to this topic

[Hongwu Shen]

I was trying to simulate plasma concentrations at specified timepoints for three subjects who received different doses, but the results showed identical concentrations for the three IDs. here is how I did the simulation.

 

I had done a simulation for the same senario of three subjects, but with sequencial timepoints (simulate every one hour for 2040 hours). I assume no subject variability, so the only difference is dose. I froze the PK parameters, In the Run option, I chose the Naive-pooled method, and SIM/Pred check run option, and set N=0. In the "tables" of preditive check options, entered "seq(0,2040.1) for times and C for variables. This one gave me the results I wanted. three subjects had different concentrations.

then I copyed this simulation and pasted. for the new one, the only thing changed was that the "seq(0,2040.1) changed to 23 timepoints from 0 to 2040 (0, 0.5, 2, 8, ~~~~, 2040). then the simulation results gave me the same concentrations for three subjects receiving different doses.

 

I wonder anyone know why I get that results and how can I fix it. Thanks!

[serge guzy]

Dear Hongwu.

Can you send please the project. That would be the best way to debug.

best Regards;

Serge

[serge guzy]

Dear hongwu

Can you just change the engine from naive pool to FOCE ELS? In the upcoming version, the engine of choice will not be independent but the current version computes some log-likelihood and therefore engine has somehting to do with the simulation procedure.

now if you still get 3 times the same response, look at your data set that was used for simulation and make sure you have 3 patients and that you map the patient with ID. Let me know what is going on. If still problems, please send me the project.

best

serge