Biology Reference
In-Depth Information
and GON are currently being equipped with an XML based file format exchange. We have developed
an environment to extract data of metabolic networks from KEGG, BRENDA and RegulonDB and
transform them into XML based files that can be used by PNK and Renew to display the Petri net models
automatically [Chen, 2002; Chen
et al.
, 2002b].
Data preparation
The main feature of metabolic processes is that the concentration of metabolites will influence the
reaction activity of bioprocesses. Therefore, the actual concentration of any metabolite is an important
component of the model. Although some data nowadays are available to public via the Internet and other
sources, some other data may be not complete. It requires a time consumption searching throughout the
literature. In case of the unavailability of concentration values, additional experiments might be required.
However, some computational prediction and experiential data also might help. Assignment of initial
value of places is made after data gathering.
Determination of the kinetics
A series of predefined types of kinetics which are frequently used in biochemical reaction models are
collected by SBML [Hucka
et al.
, 2001]. However there are some circumstances in which the kinetic
types are not yet defined. Then the kinetic modeling strategy is to be applied to ouline the kinetic
properties of each reaction. All relationships and influences of metabolites are to be fixed by introducing
the conresponding variables into the self-defined functions.
Stoichiometry check
Many metabolic pathway schemes contain mass conservation relations that must be taken into account
in order to carry out the simulation. To check the mass conservation relations of a model we can go
to the original reaction data from databases or the literature, or calculate them manually. In fact, we
construct the model with the identification of reaction stoichiometry. Otherwise, it will lose something
when the simulation is carried out because in continuous Petri nets, the weight of arcs is disable, so that
all components involved in the reaction are changed with same rate which is defined by the transitions
function. However, in the reaction 2*A
+
B
C, the change of A should be twice than that of reaction
rate. In VON++, unfortunately, we have to add more transition from A with the same function in order
to obey the mass conservation law.
Parameter tuning and simulation
To build a model precisely requires as more as possible the variables and parameters involved in
the metabolic network. The values of variables and parameters are determined either by experimental
methods or deduced from other related values. However, it is impossible or sometimes unnecessary to
put all variables and parameters into a model. The model is plausible when main influences are included.
On the other hand, because of different purposes and situations, most data from laboratory do not fit the
model very well and vice versa. We have to compare and tune the differences in order to find suitable
ones. Then the effects of various parameters on the gene regulated metabolic networks and their relations
can be understood. The key enzymes/proteins and intermediates related in the metabolic pathway can
be determined, which will provide the necessary information to solve the metabolic bottlenecks.
