updated October 28, 2007
(recovered from archive.org after a failed disk, site moved)

Kinetics for Artificial Gene Networks

In order to properly simulate gene expression dynamics, it is important to use reallistic kinetics for each transcription step. Much is already known about the kinetics of transcription at a microscopic level, i.e. for the steps of initiation, elongation and termination. However, those kinetics are too detailed to be useful for large scale models (of more than 50 genes, possibly thousands). It is then important to select phenomenological kinetic types that reflect the fundamental molecular properties of transcription.

In addition to the requirement of reallistic kinetics, it is also important for this work to have kinetic types that are scalable. By this we mean that the software must be able to define all kinetic types in a model automatically, while still retaining a good level of realism.

The main characteristics of the kinetic types used here are:

Nucleotide concentrations are not taken into account, which effectively means that they are considered constant in the models. This should only be a problem when a gene network is considered in conditions of nucleotide starvation or high turnover.
In the absence of any inhibitors or activators, the transcription of each gene is considered to be at a basal, constant, level. However, this basal transcription rate could be set to zero.
The rate of transcription must have an asymptotic maximum. This is due to the fact that there is a finite number of copies for each gene, and transcription is catalyzed by a limited number of transcription complexes.
The effect of modifiers of transcription, activators and inhibitors, must also have saturation effects. This is due to there being only a finite number of binding sites for these molecules.
Inhibitors and activators should be able to act cooperatively, i.e. with sigmoidal activation or inhibition kinetics
To simplify the network generation, each modification effect is considered to act independently. This is seen as a limitation and it is our intention to relax this constraint in due time.

The kinetis used by our models follow one of the two generic types below. Unless stated otherwise, the data presented in this site follows Eq. 1.

Eq. 1

Eq. 2

The parameters are:

V: basal rate
I: inhibitor concentration
Ki: inhibition constant
ni: inhibitor's degree of cooperativity
A: activator's concentration
Ka: activation constant
na: activator's degree of cooperativity

The transcription rate for each gene have an arbitrary number of inhibitors and activators, which are other genes. This number is governed by the nature of the gene network topology used.

Thanks to the WayBackMachine at the Internet Archive for recovery of this page after a crashed RAID array!