Navigating biological complexity: application of design of experiments methodology to metabolic engineering and synthetic gene networks.

Biological engineering presents an attractive route to tackling many of the world’s problems, for example, the production of fuels and chemicals from renewable sources. Biological systems, however, are highly complex interacting systems containing many non-linear responses. Even within the best-studied biological systems many of the component parts remain to be revealed or their functions fully characterised. This presents a challenging proposition for the biological engineer. Our research group is interested in this challenge, in particular in the application of statistical and experimental methods that facilitate forward engineering of biological systems. Here, I will first describe the construction of metabolic pathways for the biosynthesis of petroleum-replica biofuels - these molecules are structurally and chemically identical to the fuels they seek to replace. Our metabolic engineering strategy, involving the introduction of up to 10-genes from four organisms in to Escherichia coli, resulted in the production of a range of hydrocarbons that are not naturally biosynthesised and that correspond to the range of molecules found in retail diesel. I will discuss the difficulties of optimising systems that encompass multiple genetic and environmental factors, expensive-to-assay responses and multiple objectives. I will introduce some of the concepts of statistical Design of Experiments - a search strategy that permits a highly efficient, multifactorial search of the biological ‘design space’. The methodology is intuitive for experimental biologists, and importantly it provides statistically structured data sets that enable predictive modelling. Finally, I will conclude by discussing some of our current research applying this methodology to the engineering of cell free gene networks and cell free biochemical pathways.