UC San Diego

Microbes are increasingly engineered for a large variety of valuable applications. Designing microbial strains remains challenging due to the complexity and knowledge gaps with biological systems. Bioengineers have a unique advantage over other engineering disciplines: biological systems have been solving challenges long before human intervention through adaptive evolution. Adaptive Laboratory Evolution (ALE) methods leverage this natural problem-solving process to elucidate biological functions and generate application solutions. The promise of ALE methods has resulted in the generation of a substantial amount of public ALE data, which in aggregate, could contribute new insights towards ALE-derived strain design. The work of this dissertation combined aggregated public ALE data and rich mutation annotations to design strain variants with possible value in applications. Public ALE mutational data was consolidated into a web-accessible database that can report and export the aggregated ALE data. ALE metadata was used to statistically associate mutations to experimental conditions, reducing the amount of conditions to consider during mutation functional analysis and therefore deconvoluting the context of adaptive mutations. Multi-scale and multi-dimensional mutation annotations were used to identify the mutation effects, structural features, regulatory features, and cellular subsystems that ALE mutations converged upon. Meta-analysis of the aggregated ALE data revealed the principles underlying adaptive mutation trends, which were then used to design novel strain variants. The results of this dissertation demonstrate how strain design principles can be extracted from aggregated ALE data to enhance microbial engineering efforts.