Multi-objective optimization of metabolic networks for intracellular flux prediction. Application and analysis in cyanobacteria Synechocystis

Programa de apoyo I+D+I UPV. 2011

Synthetic biology is a new field of research that combines science and engineering. Its objective is the design and construction of novel biological systems for a specific purpose. There are four areas of research in this discipline (biology, chemistry, engineering, and computer science) that represent the different approaches taken by researchers.
From the point of view of engineering and computer science, engineers see biology as a field rather than a technology. Objectives are principally the design and construction of biosystems that manipulate chemicals, make materials, produce energy, etc. Considerable effort is being dedicated to developing tools to facilitate and cheapen the design of these biological systems.
A key factor in synthetic biology is understanding the behaviour of biological systems and their responses. The first step towards a model of this behaviour is describing the biochemical reactions inside the biosystem. It is essential to obtain a modular description of the various parts and mechanisms that govern the system (for example, modelling facilitates an analysis of how a system responds when it forms part of a new device).
The intracellular fluxes that occur in biological systems in silico can be calculated under the assumption that these systems work optimally with respect to certain biological criteria. The capabilities and flow distributions of a biochemical network have been predicted using, for example, metabolic flux balance analysis (FBA). The FBA method only requires a stoichiometric model of a network. However, as a linear system of equations describing the mass balance at a steady state is generally indefinitely compatible, it is necessary to define appropriate criteria and constraints to find a single solution.
Various interesting questions regarding the principles for the optimal functioning of biological systems have emerged. For example: what are the objectives or criteria that these systems use to present an optimal behaviour? Most investigators believe that this could be the maximisation of growth (biomass yield). Other researchers, however, point to the maximisation of the energy the system can produce (ATP yield), or the minimisation of intracellular fluxes.
A more realistic approach to biological system performance is to consider the simultaneous optimisation of two or more targets, often in conflict. This is termed multi-objective optimisation, where, in contrast to the case of a single objective, there is no single solution that optimises all the objectives simultaneously. Rather, the solution is a set of solutions, where each is optimal in some sense and none is better than the others. This set of solutions is termed the Pareto set and is produced by algorithms or multi-objective optimisation techniques.
Preliminary research indicates that this optimisation methodology is capable of responding to problems arising with complex metabolic models. Evidence of the novelty of the research outlined here are the relatively few results found in the literature when compared to related areas.
The project poses the challenge of whether intracellular fluxes can be correctly estimated by simultaneously considering various criteria for optimal performance under the working hypothesis that the optimal distribution of these flows is, in reality, governed by a single cell approach (as for example, the tendency of micro-organisms to grow at their maximum rate), while at the same time, the biosystem operates according to various simultaneous criteria or targets.
In this sense, the research team combines experience and ‘know-how’ that is likely intersect in the line of work. Researchers from the InterTech group have extensive experience in the field of synthetic biology and have received international recognition for their work; while researchers from the Institute of Control Systems and Industrial Computing have extensive experience in developing algorithm for optimisations and decision making based on evolutionary techniques and artificial intelligence.