Boyan Yordanov, Neil Dalchau, Paul Grant, Jim Haseloff, Stephen Emmott, and Andrew Phillips
Recent DNA assembly methods have enabled the physical construction of large-scale biological devices but designing systems with a given behavior remains a challenge in synthetic biology. To enable the engineering of complex and robust systems, integrated experimental and computational methods are needed that allow precise experimental data to be collected and used to construct reliable computational models, that serve to make accurate predictions, leading to the design of functionally correct devices.
While fluorescent reporter proteins provide a convenient experimental tool for characterizing gene expression in vivo, such measurements often depend on experimental conditions. The use of relative measurements based on a reference (standard) promoter has been proposed as a strategy for attenuating such variability [Kelly et al., 2009]. Recently, an alternative experimental and mathematical characterization framework was proposed by J. Brown in [Brown, 2011]. There, the use of spectrally-distinct reporter proteins led to a ratiometric characterization protocol where an internal reference signal provided more direct measurements of relative promoter activity. This data was also integrated in a mathematical framework extending [Leveau and Lindow, 2001], which allowed the parametrization of gene expression and regulation models. As the cost of characterization decreases and more data becomes available, computational tools capable of handling such information are required as part of biodesign platforms.
In this abstract, we present our recent results on extending, implementing and applying the ratiometric characterization methods from [Brown, 2011]. Specifically, we developed alternative strategies for representing and processing experimental data, and incorporated a Bayesian characterization framework to allow the parametrization of mechanistic mathematical models that capture uncertainty. This allowed us to compare competing hypothesis about the influence of experimental conditions and the precise mechanisms involved in gene regulation. We implemented these procedures within the GEC environment [Pedersen and Phillips, 2009], allowing the method to be used conveniently as part of a computer-aided design process in synthetic biology [Dalchau et al., 2012].
In International Workshop on Biodesign Automation