Abstract

This work presents an advanced modelling procedure, which applies both structural
modelling and kinetic modelling approaches to the trypanothione metabolic
network in the bloodstream form of Trypanosoma brucei, the parasite responsible
for African Sleeping sickness. Trypanothione has previously been identified as
an essential compound for parasitic protozoa, however the underlying metabolic
processes are poorly understood. Structural modelling allows the study of the
network metabolism in the absence of sufficient quantitative information of target
enzymes. Using this approach we examine the essential features associated
with the control and regulation of intracellular trypanothione level. The first
detailed kinetic model of the trypanothione metabolic network is developed,
based on a critical review of the relevant scientific papers. Kinetic modelling
of the network focuses on understanding the effect of anti-trypanosomal drug
DFMO and examining other enzymes as potential targets for anti-trypanosomal
chemotherapy.

We also consider the inverse problem of parameter estimation when the system
is defined with non-linear differential equations. The performance of a
recently developed population-based PSwarm algorithm that has not yet been
widely applied to biological problems is investigated and the problem of parameter
estimation under conditions such as experimental noise and lack of information
content is illustrated using the ERK signalling pathway. We propose
a novel multi-objective optimization algorithm (MoPSwarm) for the validation
of perturbation-based models of biological systems, and perform a comparative
study to determine the factors crucial to the performance of the algorithm. By simultaneously
taking several, possibly conflicting aspects into account, the problem
of parameter estimation arising from non-informative experimental measurements
can be successfully overcome. The reliability and efficiency of MoPSwarm
is also tested using the ERK signalling pathway and demonstrated in model validation
of the polyamine biosynthetic pathway of the trypanothione network.
It is frequently a problem that models of biological systems are based on a
relatively small amount of experimental information and that extensive in vivo
observations are rarely available. To address this problem, we propose a new
and generic methodological framework guided by the principles of Systems Biology.
The proposed methodology integrates concepts from mathematical modelling
and system identification to enable physical insights about the system to
be accounted for in the modelling procedure. The framework takes advantage
of module-based representation and employs PSwarm and our proposed multiobjective
optimization algorithm as the core of this framework. The methodological
framework is employed in the study of the trypanothione metabolic network,
specifically, the validation of the model of the polyamine biosynthetic pathway.
Good agreements with several existing data sets are obtained and new predictions
about enzyme kinetics and regulatory mechanisms are generated, which
could be tested by in vivo approaches.