Coping with complexity: Machine learning optimization of cell‐free protein synthesis

Abstract

Biological systems contain complex metabolic pathways with many nonlinearities and synergies that make them difficult to predict from first principles. Protein synthesis is a canonical example of such a pathway. Here we show how cell-free protein synthesis may be improved through a series of iterated high-throughput experiments guided by a machine-learning algorithm implementing a form of evolutionary design of experiments (Evo-DoE). The algorithm predicts fruitful experiments from statistical models of the previous experimental results, combined with stochastic exploration of the experimental space. The desired experimental response, or evolutionary fitness, was defined as the yield of the target product, and new experimental conditions were discovered to have ∼ 350% greater yield than the standard. An analysis of the best experimental conditions discovered indicates that there are two distinct classes of kinetics, thus showing how our evolutionary design of experiments is capable of significant innovation, as well as gradual improvement.

A cost-effective polyphosphate-based metabolism fuels an all E. coli cell-free expression system

“…Cell-free expression systems fueled by nonphosphorylated glycolytic intermediates (maltodextrin, glutamate, glucose and pyruvate) are on average about 200 times more efficient than cell-free expression energized by expensive highenergy phosphate donor molecules (fructose-1,6-biphosphate, CP, 3-PGA and PEP). Moreover, the efficiency of the system presented herein, could be further improved by replacing expensive nucleoside triphosphates (NTPs) with nucleoside monophosphate (NMPs) and exploit the endogenous enzymes to regenerate NTPs (Calhoun and Swartz, 2005;Jewett et al, 2008) The novel ATP-regeneration system presented in this work is suitable for semi-and continuous systems for in vitro protein synthesis (Spirin and Swartz, 2008), industrial applications (Swartz, 2006), high-throughput experiments (Caschera et al, 2011) and large-scale reaction using cell-free or enzyme technology (Butler, 1977). Our study also demonstrates that cell-free transcription-translation systems are valuable platforms for understanding and developing novel metabolic pathways (Zhang et al, 2007;Zhu et al, 2013).…”

Protocols for Implementing an <em>Escherichia coli</em> Based TX-TL Cell-Free Expression System for Synthetic Biology

“…For pure increases of yield, optimization can be done by machine-learning approaches. 28 Questions of resource competition and limitation can be addressed by mathematical models verified using experimental data.…”

Experimental Design for Combinatorial Experiments

“…The optimum space was complex and most easily visualized using an Excel heat map to display the region. The method has been extended to optimization of protein synthesis kits (Caschera et al, 2011a) and evolvable artificial cells (Caschera et al, 2011b).…”