I should point out that although I may have told John that bimolecular
reactions might pose problems for the 'universal method' it wasn't I who
first pointed this out, I can't remember who it was that told me (there
are a number of possibilities) but it was quite some time ago now.
The ideal of the 'universial method' has been known for many years
particularly since 1985 when the branch theorems were first published.
In addition a systematic method was immediately apparent when Reder
published here paper in J theor biol, namely to simply work out
the null space from the stoichiometry matrix. However, what might not
be so well known is a subtle modification to the systematic method
first given to me by John Barrett (and implemented in my program imap)
which is to selectively remove columns from the stoichiometry matrix
in order to achieve different patterns of gene modifications and
specific patterns of flux adjustments, essentially different branching
theorems except with the ability to be more specific about what
movements in flux one might desire. Unfortunately the 'universial
method' does not appear to be so universal especially when confronted
with reaction steps that are biomolecular which means that far more
enzyme concentrations have to be modified than would otherwise be
necessary.
Anyway, who every first thought about the biomolecular problem should
give us a run down on its implications.
Herbert Sauro
