GGAssembler: Precise and economical design and synthesis of combinatorial mutation libraries

Abstract

Golden Gate assembly (GGA) can seamlessly generate full‐length genes from DNA fragments. In principle, GGA could be used to design combinatorial mutation libraries for protein engineering, but creating accurate, complex, and cost‐effective libraries has been challenging. We present GGAssembler, a graph‐theoretical method for economical design of DNA fragments that assemble a combinatorial library that encodes any desired diversity. We used GGAssembler for one‐pot in vitro assembly of camelid antibody libraries comprising >105 variants with DNA costs <0.007$ per variant and dropping significantly with increased library complexity. >93% of the desired variants were present in the assembly product and >99% were represented within the expected order of magnitude as verified by deep sequencing. The GGAssembler workflow is, therefore, an accurate approach for generating complex variant libraries that may drastically reduce costs and accelerate discovery and optimization of antibodies, enzymes and other proteins. The workflow is accessible through a Google Colab notebook at https://github.com/Fleishman-Lab/GGAssembler.