Paul Schlosser wrote:
>>I would suggest that there are two primary reasons for the initial
>development of matrix tools:
>>1) in the case where one does not have equations for all the enzymes
> involved, or you want to quantify the behavior of a system with
> too many variables to develop a comprehensive model, they allow one
> to analyze the system without building a formal model;
>2) a historic lack of the compute-power and software needed to do the
> fully dynamic stuff.
>If I can build a simulation model and put in everything I know about the
>enzyme kinetics, over large concentration ranges, then any subsequent
>matrix approach will, at best, extract only a subset of the information
>in the model. So why settle for a subset? I can do local sensitivity
>analysis (e.g, steady-state CCs), finite-difference sensitivity analyses,
>and dynamic sensitivity analyses, on models of any complexity or level(s)
>of organization, for every single parameter in the model, without deriving
>a separate method for each type of problem.
>
I think Paul is about right in the first part, though reason 2 has not been
a very significant factor for some time now.
However, when he says 'If .....' of course his argument is pretty
unassailable provided his conditions are met. But:
1. Michael Savageau pointed out (in 1976 I think) that the enzyme glutamine
synthase is known to interact with about 15 metabolites (substrates,
products, effectors). There is the possibility that all of these interact
with one another. If you are going to have a kinetic equation that
characterizes the enzyme at any combination of concentrations of these, you
have got to come up with a suitable equation and then fit the parameters to
it on the basis of in vitro experiments. Well, let's assume 'large
concentration ranges' can be represented by a set of 4 suitably chosen ones,
and that to look for all possible interactions all combinations of
concentrations of the 15 metabolites are investigated, that's 4 to the power
15 rate measurements to be made, i.e. 10 to the power 9 and we've got one
enzyme sorted. Until someone discovers effector number 16 ............
That was the basis on which Savageau doubted we would ever have sufficiently
detailed kinetic information to make Paul's 'if' a starter. From my own
experience with a more limited problem (oxygen binding to haemoglobin), I
think Savageau was then, and still is now, right to be sceptical. I would
not conclude that we have too little information to ever justify simulation,
but I do not think we have enough to be confident we can dispense with all
other ways of looking at a metabolic problem. We'll have to look at any
system from a number of angles and decide whether the different views are
consistent or not.
2. There are too many examples of perfectly respectable in vitro kinetic
data not leading to sensible answers when used to calculate in vivo
behaviour (for reasons that can often only be speculated about) to be
confident that even if we could get round Savageau's objection we would have
solved all the problems of doing accurate simulation. For example, Barbara
Wright has had a programme running for years on slime mould metabolism that
has combined kinetic measurements and computer simulations, and she always
has to change the measured parameters to make the computer simulations work.
Even at the level of single enzymes there are discrepancies. Calculated
activities do not match in vivo fluxes. Aldolase in glycolysis is a
persistent offender (e.g. Dan Frankel's experiments on metabolite changes on
over-expressing aldolase).
I don't want to appear pessimistic. I believe we should use available
biochemical knowledge to construct simulations exactly so that we are
confronted by the gaps in our knowledge. At least in this way we are making
a test of the quality of our understanding that is simply being ducked if we
only use qualitative approaches. If there are no gaps in our knowledge,
then we can give up using control analysis because, as Paul says, we can
find all the answers from the model. But we've got a long way to go, and
the prospects are not improving now that so little enzymology gets funded
because the cloners have persuaded people that the answers can all be found
in the gene sequences.
And I favour moderation of the news group, but I don't want to read any more
arguments about it.
David
=========================================================
David Fell,
School of Biological & Molecular Sciences,
Oxford Brookes University,
Oxford OX3 0BP, Tel. +44 (0)1865 483247
UK FAX +44 (0)1865 484017
daf at bms.brookes.ac.uk
See my book at: http://www.portlandpress.co.uk/books/isbn/185578047X.htm
