Simulation makes no sense

I have some more troubles with Chance1960_Glycolysis_Respiration model. I have set the glucose's trigger concentration to 0.3. When I run it for 1000 s, I get the step increase of glucose and mildly decreases in time, while oxygen, lactate or pyruvate stay at near 0 levels:

When I run it for 3600 s, I get an error, while the graph remains the same:

Conversely, when I run it for 1 hour which is the equivalent, results are different.

When I run it for 2 days, the step increase in glucose that was seen in the previously when run at shorter periods of time, is no longer seen, and glucose level is flat from the very beginning at near 0:

I know these profiles are not logical and conflicting, which points to me that the simulation is obviously wrong. How could I fix that? Thanks very much!

Juan Carlos

[Juan Carlos Cao Yao]

By the way, my model comes from BioModels database's curated files.

Answer 1

[Jeremy Huard]

Hi Juan Carlos,

as Pax suggested in his answer to your previous question, this model does not contain units, so you can't make use of the unit conversion feature in SimBiology. Which also means that changing the unit of the StopTime won't have any effect.

This StopTime will always be in the unit chosen by the model author to define all quantities, which is here 'second'.

If you want to simulate for 1 hour, you will have to set StopTime to 3600 sec as you previously did.

Now, the error you get suggests that the integration tolerances are not met. You can try to decrease them and see it runs.

You can modify the integration tolerances in the simulation settings:

Answer 2

[Jeremy Huard]

By the way, I think had changed the glucose injection to 0.3 instead of 0.003 mainly to make easier to spot the event.

But if you revert it to 0.003, the model runs and simulate 3600 sec with the default tolerances without issues.

Answer 3

[Juan Carlos Cao Yao]

Wow, Jeremy, thank you for this amazing skills showcase. It understand my problem perfectly now and solved it. Integration tolerance is something deep about mathematics, but I get that depending on tolerance, there is a limit to the amount of substance the software can handle, say glucose. Thank you very much.