>This is a seducing idea, but does it have enough theoretical background? For
>instance, how can one distinguish between effects caused by channelling and
>those due to inhibitors or activators with variable concentrations?
In broad terms, "the theory" makes statements of the kind: "Given that a
metabolic pathway has properties 'X' and 'Y', we can conclude that it
must also have property 'Z'."
You can always find examples of pathways that do not fit the theory; i.e.,
that do not have properties 'X' and 'Y'. This does not make the theory
invalid or mean that it is not useful. It just means that this is not a
universal theory of all metabolic processes. However "the theory" is not
some fixed thing that has once been proven leaving only its application.
MCA can and indeed does continue to be extended to address more and more
complex problems.
Even given that one is restricting one's attention to the "unrealistic"
scenario of assuming "X" and "Y" (when these may not be true in many
cases), a primary value of MCA is that the conclusion, "Z", is often not
obvious and goes against the 'classical', 'intuitive' 'understanding' of
metabolic pathway behavior that many non-mathematical experimentalists
have had. In short it tells us that some of the classical notions which
were developed in the absence of rigorous mathematics were simply incorrect.
Finally, to quote a colleague of mine, "a model is most useful when it
fails". In this context, if one finds a pathway in which property "Z" is
can be shown not to hold, then one can conclude that either "X" or "Y"
must not be true. If "X" and "Y" are considered to be hypotheses about
the pathway, then one has disproven a hyptheses, something that is very
useful in science.
> I suspect that Paul Schlosser's "Black Frame" analysis deals with this
>kind of problems. He is going to send it to me and then I will be able to
>comment competently.
Actually, the "Black Frame" does not deal with channeling. In my work I
also assumed that each step within the frame (the "transparent" area that's
being explicitly modeled) is catalyzed by a separate enzyme, that the levels
of the enzymes vary independently, and that the steps are linked by pools
of free metabolites. You could amend mine to deal with those, but I have
not done so. (I'm working in toxicology now, and that takes all my time.)
The drawback to my approach is that it requires some very delicate and
exacting experimental work and is limited by the methods available for
modulating a pathway and measuring the resulting changes in levels of
activators, inhibitors, etc.
I guess I better send something out today.
-Paul
