Fluxes Optimization with Evolutionary Algorithms

[V0idLord]

Hello,

I am working on a task using a kinetic SBML model to optimize the flux of a reaction. I used basico package in order to add my reactions and parameters to the model, but got stuck on the optimization task itself, specifically on how to set up the objective function, I am sorry to bother you if this isn't the right place for these questions, I read the optimization notebook in docs but I wasn't able to solve the problem. Specifically, how to write the expression in set_objective_function(expression='Values[f].InitialValue', maximize=True) in order to maximize a flux?

How exactly do I set a flux as the maximization objective function and how to create the list of targets for optimization (Ki, Km and Vmax in my case)?
I want to implement this as a script, as such I prefer doing it in basico instead of the GUI COPASI.

Any help is appreciated :)

[fbergmann]

This is the right place to ask these questions. To maximise the flux of a Reaction say J1 the expression would be (J1).Flux and then you'd add the bounds on ki, km and vmax as parameters to vary in the optimisation. I'll try and put together an example for that early next week.

[V0idLord]

Thank you, I was able to get it running.

[V0idLord]

One more question, is it possible to limit the number of mutations when optimizing metabolic fluxes? For example, finding solutions where only 5 or less parameters were modified? From the results I got from my testing run it looks like a lot more parameters are altered, but while solutions obtained have a better fit, they would be impractical to implement in a strain.

[fbergmann]

There is no way to specify that for the optimization algorithms. In that case it would be best to run an optimization with only those parameters that would be feasible to do. Or first run a sensitivity analysis for the parameters you selected, to see which one have more impact on the desired flux, and then only select those.