Simulation results do not comply with anticipated results

[cel.c...@gmail.com]

Hello everyone,

I am new to COPASI and I started with a small Model (which I am aiming to expand), which describes simple Receptor-Ligand interaction. 

I set the Species concentrations and Reactions as seen in the figure below and I use the Hybrid(Runge-Kutta) as solver.

In the results I see that the Ligand decreasing over time as expected, whilst the number of bound receptors (MOR) and the number of Ligand-Receptor compounds (MOR:L) do not change over time.

Does anyone know what the problem is? I really do not understand the issue. 

Thanks in advance

[sven....@gmail.com]

Hello, 

it would be easier to be sure what the problem is if you attach the model, but the first thing that comes to mind is the actual numbers in the plot: The concentration for L is 10^8 times larger than the Receptor concentration, so even if it changes, you would not see anything in the plot. 

If you look at the receptor concentrations separately and still do not see changes, it could be the time scales. It is possible the receptor concentrations change so quickly that all changes happen before the first point plotted.

Sven

[cel.c...@gmail.com]

Hello Sven,

(I hope I didn't respond to your response multiple times now)

I did check the output for the other species without the Ligand and the particle numbers over time didnt change. Please also find attached the model file.

Thank you in advance

[Hoops, Stefan (sh9cq)]

Hello ???

Your binding reaction "Ligand binding" has a flux of:

1.5882832785960066 * 10^-12

at time 0, which means that you would need approximately 10^12 minutes
to see a change of 1 in the numbers of MOR or MOR:L.

Please check the rate constants in your reactions.

Thanks,
Stefan

--
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

[sven....@gmail.com]

Hello Cel, 

are you planning on doing a stochastic simulation? If not, I would recommend to stay with the LSODA solver for now. With LSODA you will see a (tiny) change in the bound receptor. It is extremely small, because the concentration of 100 molecules of receptor in one litre is tiny, resulting in a correspondingly tiny reaction rate for the binding (because the unit of the rate constant is in terms of concentration, not in terms of particle number).

The behaviour of the system is completely dominated by the degradation of L, while the binding barely happens. 

In a deterministic simulation (with LSODA) you can at least see the extremely slow binding, in a stochastic (or hybrid) simulation the binding would happen so rarely that you would never encounter a single binding event. 

Sven

[sven....@gmail.com]

One way to start fixing the model would be to use a realistic volume (assuming you are aiming for realistic molecule counts). Otherwise a realistic count of receptor molecules (like 100) would lead to microscopic concentrations, which would then require very unrealistic rate constants to compensate.

Sven

[cel.c...@gmail.com]

Hello Sven,

thank you very much, this helped me now I have a clue on how to debug.

Best!