The differences between pFBA and normal FBA

[complicatedZ]

Dear developers,
Compared with normal FBA, we know that pFBA minimizes the total sum of flux at the same time. And as I calculate the total some of flux optimized in two different ways, the results are really close as below.

I wonder it's a coincidence or the normal FBA optimization designed to minimize the sum of flux to some extent, because due to many loops existed in the model, the sum of flux could be very high if you just want to maxmize the growth rate.
Best Regards.

[Midnighter]

Is that the small textbook model or a genome-scale model? That may have something to do with it.

Additionally, there is probably solution caching going on. You first compute the pFBA which uses the optimal value from a normal FBA, sets that as a constraint, then minimizes the fluxes. So when you next perform a normal FBA, the solver may detect that the current state is already an optimal solution and does not change the flux distribution at all. It's possible that you will see a larger difference when you first do a normal FBA and then a pFBA.

[complicatedZ]

Thank you for your reply! By the way, the model is a genome-scale model including more than 2,000 reactions, is there something special?