Assemblatron: An Automated Workflow for High-Throughput Assembly of Big-DNA Libraries.

The ability to synthesize 50-kb+ DNA molecules has tremendous potential in the fields of genome engineering, metabolic engineering, and synthetic regulatory genomics. Despite tremendous achievements in these fields, such as the completion of the first synthetic eukaryotic genome, assembling custom big DNAs remains slow, expensive, and laborious. In this work, we present a set of improvements to yeast-based DNA assembly methods that enable medium-to high-throughput big DNA experiments. In particular, we i) developed two easy-to-use vector systems: Jack In the Box (JIB), which deploys a split marker and split centromere strategy to reduce background, and Selection of URS Recombinational Excision (SURE) which exploits URS silencer removal from the marker. These strategies sharply decrease the number of colonies containing an empty vector, greatly reducing the amount of screening required to find correct clones. ii) We improved yeast transformation efficiency, increasing the likelihood that all required DNA segments are co-transformed into the same cell. iii) We developed an automation pipeline for segment mixing through single-colony isolation. We also established three phenotype-driven assays that directly and rapidly report the frequency of correctly assembled large DNA molecules. Harnessing these improvements, we show a high success rate in assembling diverse big DNA libraries by combinatorial assembly and oligo-guided architecture in yeast (OLIGARCHY), a novel method that allows cost-effective reuse of segments for distinct assemblies with reduced effort and cost. Using OLIGARCHY, we designed 96 structural variants of a ∼73kb construct, and a single person was able to correctly assemble 90 of these in 7 days, for a total of 7 Mb of assembled DNA. Finally, we used the SURE vector to greatly improve existing methods to turn oligonucleotides into gene-sized double-stranded DNA segments, efficiently turning 80 overlapping 60-mers into ∼3-kb DNAs directly, without PCR amplification.