Computer Analysis of Steady State Models of Enzyme Networks
77
system and a sequence of commands taken from a permitted list. The more
sophisticated user is still quite free to exploit the full power of a high level
language.
If it is desired to alter the structure of the system under study this will
mean making adjustments to A or even, if a new system is to be considered,
replacing A entirely. Altering parameters and asking "questions" about the
system is done in C and C will be completely replaced at each simulation run.
Finally, the provision of new "questions" for use in C will mean adding
routines to section B.
At no time therefore is there a fixed program which merely accepts
data. This leaves us with the considerable task of altering an extensive program quite frequently in sections A, B, and C.
To solve this problem a "pre-compiler" was written in collaboration
with Dr. J. G. BURNS l . This is essentially a program for merging, in a very
flexible way, text received from paper tape with text already held on magnetic tape in various files. At present the "pre-compiler" maintains seven
program files and a library and has a set of commands enabling the user to
modify the contents of one or more files, to build up programs from parts
held in different files, and to present the programs thus generated, sequentially to the compiler. All these operations being on a "load and go" basis,
the complete set of operations requires only one paper tape as input and
thereafter proceeds automatically.
A person without "computer know-how" can easily combine a system
A with a set of analytic operators contained in B, about which he need
know very little, and then ask "questions" in a flexible way in part C.
Furthermore, systems previously studied can be left on file and again placed
in A when required, similarly, one of several alternatives could be chosen
for B.
The cost of the pre-compiler phase is about 30 seconds overhead and
15 seconds for each simulation assembled and presented to the compiler,
which is trivial compared with time spent in compilation: 3 to 4 min
and in actually running the simulation: 12 min or more.
Type of Metabolic System
The computer methods so far discussed impose no limitation on the
model but in what follows the model is limited to be that of enzymes and
their substrates in solution in a well stirred space.
Since our primary interest is in the properties of the steady state, it is
advantageous to make no mention of enzyme-substrate complexes but
to use "net-flux" expressions similar to those described by CLELAND [2]. A
I Dr. J. G. BURNS, Dept. of Natural Philosophy, Univ. of Edinb.
77
system and a sequence of commands taken from a permitted list. The more
sophisticated user is still quite free to exploit the full power of a high level
language.
If it is desired to alter the structure of the system under study this will
mean making adjustments to A or even, if a new system is to be considered,
replacing A entirely. Altering parameters and asking "questions" about the
system is done in C and C will be completely replaced at each simulation run.
Finally, the provision of new "questions" for use in C will mean adding
routines to section B.
At no time therefore is there a fixed program which merely accepts
data. This leaves us with the considerable task of altering an extensive program quite frequently in sections A, B, and C.
To solve this problem a "pre-compiler" was written in collaboration
with Dr. J. G. BURNS l . This is essentially a program for merging, in a very
flexible way, text received from paper tape with text already held on magnetic tape in various files. At present the "pre-compiler" maintains seven
program files and a library and has a set of commands enabling the user to
modify the contents of one or more files, to build up programs from parts
held in different files, and to present the programs thus generated, sequentially to the compiler. All these operations being on a "load and go" basis,
the complete set of operations requires only one paper tape as input and
thereafter proceeds automatically.
A person without "computer know-how" can easily combine a system
A with a set of analytic operators contained in B, about which he need
know very little, and then ask "questions" in a flexible way in part C.
Furthermore, systems previously studied can be left on file and again placed
in A when required, similarly, one of several alternatives could be chosen
for B.
The cost of the pre-compiler phase is about 30 seconds overhead and
15 seconds for each simulation assembled and presented to the compiler,
which is trivial compared with time spent in compilation: 3 to 4 min
and in actually running the simulation: 12 min or more.
Type of Metabolic System
The computer methods so far discussed impose no limitation on the
model but in what follows the model is limited to be that of enzymes and
their substrates in solution in a well stirred space.
Since our primary interest is in the properties of the steady state, it is
advantageous to make no mention of enzyme-substrate complexes but
to use "net-flux" expressions similar to those described by CLELAND [2]. A
I Dr. J. G. BURNS, Dept. of Natural Philosophy, Univ. of Edinb.