Towards the Automated Design of Bacterial Genomes for Tailored Applications

In the not-too-distant future, it will become far easier to design and build large genetic systems “from scratch” rather than troubleshoot the many difficulties that arise from “legacy” DNA code. To reach that inflection point, we will need radically new models and design algorithms that can reverse- and forward-engineer the function of large genetic systems according to its sequence, including its expression levels, regulatory dynamics, and metabolic fluxes.

Here, we present new models and algorithms capable of designing and optimizing the sequences of large bacterial genetic systems, from 50 kbp to 5 Mbp, to satisfy both local and global design rules governing transcriptional regulation, translation, mRNA stability, metabolite and resource utilization, biosynthesis, and evolutionary stability. Together, the design rules eliminate several undesired “legacy” genetic elements, while ensuring targeted protein expression dynamics, minimal resource utilization, and maximal evolutionary robustness. To do this, we’ve expanded our toolbox of biophysical models and developed new system-wide constraints to predict how individual genetic parts collectively control operon function, and how multiple operons collectively control resource utilization. We present experimental validation of a newly developed biophysical model that predicts the mRNA degradation dynamics within operons, including strong coupling between translation and mRNA stability. We then demonstrate our design algorithms on selected examples: natural product gene clusters, the central metabolic network of E. coli, and the small genome of JCVI-Syn3.0. The presented algorithms take us a significant step closer to the automated design of whole genomes for tailored applications.