475309 The Power of Prediction: Rapid Optimization of Metabolic Pathways without High-Throughput Screening

Wednesday, November 16, 2016: 3:30 PM
Continental 6 (Hilton San Francisco Union Square)
Howard M. Salis, De Novo DNA, University Park, PA; Department of Chemical Engineering / Biological Engineering, Pennsylvania State University, University Park, PA

The Salis Lab and De Novo DNA have developed an integrated computational-experimental pipeline that rapidly designs and optimizes many-protein genetic systems, such as metabolic pathways & networks, inside engineered micro-organisms, while requiring only a small number of characterization experiments. Our pipeline can be applied to any large genetic system to optimize its performance, using only two design-build-test-learn cycles. To do this, we've combined several predictive models, design rules, and optimization algorithms to map the genetic system's sequence-expression-activity relationship, predict its optimal protein expression levels, and prioritize protein engineering efforts. Our platform is web-accessible at http://www.denovodna.com/software, and has been used by over 6000 registered researchers to design over 100,000 synthetic DNA sequences. We present case studies to demonstrate the pipeline's process.

Extended Abstract: File Not Uploaded