How to use the optimal set?

[weix...@umn.edu]

For parameter estimation in Copasi,  if you have different initial guess, then you will have different parameter sets that fits the data. I wonder that how to decide which one to choose? And how many run you need for the optimal method you choose? Thanks! 

Best

Ning 

[Frank Bergmann]

Ideally, you would know from the experimental conditions under which the data was gathered, what your initial conditions are. If you are really unsure, you could try and specify that the initial value should also be estimated by COPASI. 

best

Frank

[Marc Lefranc]

Hi,

this is a well known problem in parameter estimation, and the larger the number of parameters, the worse the problem. The idea is that the score function has many local minima in which the procedure can be trapped. It is practically impossible to sample high-dimensional parameter spaces, making it difficult to be confident that you have the best solution.

You can try to scan on one parameter, optimizing the others, and then do the same thing with another parameter. Sometimes this takes you in a deeper valley.

Best,
Marc

--
You received this message because you are subscribed to the Google Groups "COPASI User Forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to copasi-user-fo...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/copasi-user-forum/6c4eabd3-846f-4f8d-8a9a-d59cefb7d2a7n%40googlegroups.com.

[Ning Wei]

Thanks. I wonder that how to scan on one parameter, optimizing the others? 

Best

NIng 

To view this discussion on the web visit https://groups.google.com/d/msgid/copasi-user-forum/1181d9e7-013b-10d9-266b-bd8d5a917903%40univ-lille.fr.

[Hoops, Stefan (sh9cq)]

Hello Ning,

Please use parameter scan with subtask parameter estimation or
optimization. This will give you an optimal solution for each scanned
parameter value.

Thanks,
Stefan


On Fri, 2021-03-12 at 08:52 -0500, 'Ning Wei' via COPASI User Forum

> > https://groups.google.com/d/msgid/copasi-user-forum/1181d9e7-013b-10d9-266b-bd8d5a917903%40univ-lille.fr
> > .

--
Stefan Hoops, Ph.D.
Research Associate Professor
Biocomplexity Institute & Initiative
University of Virginia
995 Research Park Boulevard
Charlottesville, VA 22911

Phone: +1 540 570 1301
Email: sho...@virginia.edu

[Ning Wei]

Hi Stefan 

Thanks. I wonder that if you have around 50 parameters to estimate, which is the best procedure to approach that problem? 

Best

Ning 

To view this discussion on the web visit https://groups.google.com/d/msgid/copasi-user-forum/1b74eee8dd45accd397366c092e225ffba1da32e.camel%40virginia.edu.