Closed EmanuelGoncalves closed 7 months ago
Hi Emanuel,
Thank you for your request. It's great that you find OptKnock solutions to problems in the genome-scale iMM904. Your problem statement is technically correct. The only thing missing is the "min_gcp" parameter. In your example, you are looking for the best solution ("sd.names.BEST"), so you can set that value to 0.0. If you use "sd.names.ANY" to find non-optimal solutions faster, you may want to put a positive value here.
Nevertheless, I now set a default value of min_gcp that is 0.0 because it makes a lot of sense. It's intuitive to start the search without a specific minimum in mind, just like you did.
On a sidenote: GLPK is probably the weakest of the four solvers, and I would usually recommend gurobi or cplex for models with 250+ reactions. More complex problems can easily run multiple days, even with stronger solvers.
I updated it in the last push 2693072293c012835bb172b190f3ea9e7f5b1a9c. Please test the updated version and if your problem persists, please reopen this issue.
Hi,
I'm testing some strain optimization using iMM904 model and managed to run OptKnock but run into some problems with OptCouple (Python 3.10.12, cobrapy 0.26.2, straindesign 1.11).
Am I missing anything?
Thank you!
Output: