Can we estimate different variance of the error model for different data set in NLME Simbiology?

1 view (last 30 days)
Hi All,
I am working on building a PBPK model in Simbiology,
Here I would like to fit my model parameters to various data, e.g. blood and tissues, using NLME method.
My question is: can we estimate the variance of the error model for each data set?
after my fitting, I can only found one estimated error model parameter and variance.
so can we get one estimated error model parameter and variance for each data set? e.g. one for blood data and one for tissue data.
Thank you!

Accepted Answer

Arthur Goldsipe
Arthur Goldsipe on 14 Feb 2022
I don't completely understand your workflow, but my guess is that this is not possible today with the NLME methods provided in MATLAB and SimBiology. The only way I know how to get separate error estimates is to do a separate NLME estimation for each tissue type.
If you can share more information, I can try to confirm. What exactly does the problem setup look like? Ideally, can you provide a sample project or maybe a paper that shows a similar problem/result? You can alway email me if you don't want to publicly share such a project.
My puzzle is what it means to perform a single NLME estimation with multiple data sets for different tissues. What does your data look like? What is your statistical model for variations in the parameters of the SimBiology model? How does the kind of tissue impact the model and estimation?
I could imagine using categorical covariates as a way to model tissue-specific variability of SimBiology model parameters. We don't support categorical covariates today, but that's something we've investigated and would consider adding support for in a future release. But it would be helpful to confirm what you would want in such a feature (especially with respect to things like standard errors or other estimates of variability).

More Answers (0)

Community Treasure Hunt

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

Start Hunting!