S.J.Eaton at newcastle.ac.uk (Simon.Eaton) writes:
About the method of Liao and Delgado, I have analysed it with John Barrett
to see if it could be applied to his (JB's) experimental system. What we've
done was to set up simulations of a small pathway (3 enzymes if I remember
well, this was more than a year ago), let it settle to a steady state, then
make a perturbation and record the transient back to the steady state. The
perturbations were of the order of 2x and 3x the original values (which is
what one can normally do in experiments). Then the data from the transients
was used to obtain the control coefficients using the methof of Delgado and
Liao. However, for the majority of cases it did not work. This seems to be due
to that method relying on linearizations. If the system is sufficiently
non-linear and the perturbation is large (surely it will be about 10% in the
best of cases) then the method seems to fail.
Now if you can find a system which is operating in a region where it is
fairly linear and you can perturb it in such a way that a) even after the
perturbation it is still in reasonably linear operation, b) you can measure
the differences between the perturbed system and the steady state, then the
method works well. Having a) and b) together might prove difficult, though.
Pedro Mendes
prm at aber.ac.uk
