Site map > MCA Info > Modelling Metabolic Processes

MetaModel: a Metabolic Modelling Program for MS-DOS

This page describes the program MetaModel for modelling the behaviour of metabolic systems in IBM PC or compatible computers.



It may seem a bit odd creating a web page now (in 1998) to describe a program for MS-DOS that was released in 1991 (Cornish-Bowden and Hofmeyr, 1991), especially now that more powerful programs such as Gepasi, SCAMP and WinSCAMP are available. Nonetheless, MetaModel may still have a role for teaching some aspects of metabolic control analysis, especially as it demands very little resources from the system on which it runs.


MetaModel started life as a program in Fortran for running on University main-frame computers and intended for teaching the principles of metabolic control to students. This was written by Athel Cornish-Bowden, but it was subsequently translated by Jannie Hofmeyr first into BBC Basic and then into Turbo Pascal, and was described under the name of METAMOD (Hofmeyr and van der Merwe, 1986). The two threads were then merged into MetaModel 2.0, described in 1991. This is the most recent version for which a description in the literature exists (Cornish-Bowden and Hofmeyr, 1991), but there is a current version, MetaModel 3.0, that incorporates a number of improvements.


Metamodel 3.0 can be downloaded from the BTK-MCA archive at Oxford Brookes University.

System requirements

MetaModel 3.0 runs under MS-DOS 2.11 or later on IBM PC-compatible machines. It requires less than 250 kbytes of memory. For practical purposes this means that it runs on any PC-type computer likely to be in use today, with memory and hardware requirements that are close to negligible by modern standards. (It will also run from the DOS command line on a modern PC running Windows).

Brief description

MetaModel 3.0 allows models to be defined that contain up to 20 reactions (enzymes) and up to 30 metabolites, and conservation constraints on the metabolite concentrations can be defined (though unlike Gepasi, for example, MetaModel 3.0 does not identify these automatically). Metabolites can have names of up to 6 characters. Once defined, a model can be stored as a file for later use, steady states and time courses can be calculated for specified initial conditions, and matrices of control coefficients and elasticities can be calculated. The program is menu-driven and some of the menus (in the form that they had in MetaModel 2.0) are illustrated in the 1991 article (Cornish-Bowden and Hofmeyr, 1991).

MetaModel 3.0 is suitable for simulating the sort of simple models likely to be considered in student classes, and it is simple enough for students with no prior exposure either to computing or to metabolic control analysis to learn enough in about 3 hours to set up models of their own and confirm the correctness of the summation relationships etc.

MetaModel 3.0 probably is not suitable for simulating complex models of the sort considered in current research projects, largely on account of the limited complexity of the kinetic equations that can be defined, but also because of its limited capacity to handle stoicheiometric constraints and processes that occur in multiple compartments. It proved incapable, for example, of simulating a model of trypanosomal glycolysis (Eisenthal and Cornish-Bowden, 1998) that was well within the capacity of Gepasi. Nonetheless, these restrictions are unlikely to be important for teaching the first principles of metabolic control analysis, and for class use it probably remains simpler to use than its more powerful brothers (except for users who cannot conceive of life outside Windows).

Time courses

MetaModel 3.0 offers a choice of the following methods for computing time courses:

For actual modelling there is little point in using the first three of these at all, but the others are included for teaching use. Unless one can see at first hand how badly the obvious method (Euler's method) performs it is not easy to be convinced that it is worthwhile bothering with more sophisticated ones.

Changes from MetaModel 2.0

The most important changes from the version described in the literature (Cornish-Bowden and Hofmeyr, 1991) are as follows:

Other possibilities

A web page written by Pedro Mendes includes brief descriptions of other metabolic modelling programs, as well as links to their web sites where these exist.