Simbiology problem

I am starting with simbiology desktop to fit my model to data. My data is imported from a .txt file. The first column corresponds to time and the 2 following columns to species concentratios at every specified time. Data imports properly and is also plotted correctly, so I think that's not a problem. I have defined a model for each species and simulates properly. I have two reactios with 4 kinetic constants, which I want to change to best fit my model to my data. In the Fit task I have green squares in all sections but when I run it gives me this error:

"The estimated object 'UnnamedReaction_1.k1' specified in the PKModelMap does not resolve to a species, parameter, or compartment object."

I am using the graphical interface sbiodesktop and I have not found much information on the web about its use. There is a point that is not clear for me and is what does mean "group" in the "dataset" tab.

Is anything I have forgot to do? I can send the files I am working with.

Thanks

Answer 1

[Arthur Goldsipe]

Hi Asterio,

I'm one of the developers on the SimBiology team. I think you may have encountered a bug that happens when a parameter is scoped to a reaction without a name. I would try to work around the bug as follows:

1. Delete k1 from the list of estimated parameters in your Fit Task.
2. Either give your reaction a name or change the scope of k1 (by right-clicking on the parameter and selecting "Change scope", so that the parameter is now scoped to the model).
3. Finally, add the parameter back to your list of estimated parameters and re-run the task.

Let us know if that takes care of your problem.

-Arthur

[Asterio]

Hi Arthur,

thank you for your answer. It did work but I still do not have a fit!

I do not really know what to do with the "group" tab.

I have in my data file one column for time, one column for biomass concentration and one column for metabolite concentration. So, for each time I have a value for biomass and another for metabolite. My model has two kinetic equations: one for biomass generation and one for metabolite generation. Two kinetic constants for each reaction, and these constants are the parameters I want to use to fit the model to my data. I have selected two responses "biomass" and "metabolite" which corresponds to my two data columns.

Which column is my "group" column? If I leave blank the group tab Matlab gives me an error "The group data column must be selected".

If I select the time column as group column I get this error message "Y must be a column vector with the same number of rows as X, which is the number of observations (N)".

I have added another column with "1" in each row and selected that as group column and get this error "OBSERVEDDATA must define target data for at least two groups."

Thank you very much.

[Arthur Goldsipe]

See my additional comments below.

Answer 2

[Arthur Goldsipe]

Hi Asterio,

Regarding your second problem with the "group" tab:

The short answer is that you should leave "group" empty, but you need to change the Estimation Method to "individual fit (NLINFIT)". Then your task should run.

It sounds like you've selected one of the other estimation methods, which can only be used when you have population or grouped data. "Grouped data" means you have multiple time courses representing different something like different conditions or replicates. You can think of each time course as a group of data.

These population methods are often used for pharmacokinetic data, where each group of data represents a different patient. The modeler then wishes to estimate "average" parameters for the population, as well as patient-specific variations.

-Arthur