Automated Design to Engineer Organisms: Scaling up Synthetic Biology to Tackle Humanity's Challenges

Abstract:
Organism engineering is the bedrock of biotechnology from producing high-value products (enzymes, materials, therapeutics) to developing cell therapies. With the latest techniques in DNA synthesis and assembly, it is now possible to construct large genetic systems with (just about) any DNA sequence of interest, enabling one to engineer sophisticated genetic systems inside cells with many genetic parts. Engineered genetic systems can act as sensors, circuits, and actuators to detect environmental states and autonomously act to change them, for example, probiotic bacteria that sense body temperatures to activate the expression of enzymes that treat metabolic diseases. However, it remains highly challenging to build such genetic systems with high-performance behaviors; there are many “tunable knobs” and inter-dependent interactions that create a “curse of dimensionality” with cryptic (unaccounted for) effects. To overcome these challenges, new approaches are needed that parallel the development of a modern engineering discipline centered around organism engineering.

Specifically, we show that it is now possible to rationally engineer genetic systems by combining predictive models of gene expression together with sequence design algorithms. Our models utilize statistical thermodynamics, kinetics, and machine learning to predict how DNA sequence controls transcription rates, translation rates, mRNA decay rates, gene regulation, and more. Leveraging these model predictions, we automate the design of genetic parts and systems (long DNA sequences) using multi-objective optimization to ensure the engineered organisms have the desired specifications (maximizing target functions, while minimizing undesired behaviors). To develop and test these models, we utilize the latest advances in oligopool synthesis, library-based cloning, and next-generation sequencing to carry out thousands of defined experiments per workflow. We illustrate our rational design approach with several recent applications, including engineering genetic systems to sense-and-respond to human biomarker proteins inside cell-free assays for medical diagnostics and engineering genetic systems to sense-and-respond to TNT inside soil systems for countermine detection.

We have also developed an interactive web-based design platform for engineering organisms, which now has over 10000 registered researchers who have designed over 900,000 genetic systems for diverse biotech applications (medical, industrial, agricultural, defense). The platform provides a “no-code” interface to our suite of predictive models & design algorithms, enabling its broad usage by the community. Altogether, these efforts demonstrate that physiochemical models can indeed predict biological functions with sufficient accuracy to automatically design genetic systems with high performance behaviors.

Bio:
Prof. Howard Salis is an Associate Professor in the Biological Engineering, Chemical Engineering, and Biomedical Engineering departments at Penn State University. He is also a member of the Bioinformatics & Genomics and Molecular, Cellular, & Integrative Biosciences graduate programs. Prof. Salis’ expertise is in the design & engineering of genetic systems in microbial organisms for diverse biotech applications (industrial, medical, agricultural, defense). His lab’s mission is to co-develop a new engineering discipline for biology through the development of predictive models & design algorithms that circumvent the need for trial-and-error experimentation. To develop and test these approaches, his lab carries out thousands of defined experiments per workflow utilizing the latest in oligopool synthesis and next-generation sequencing. Prof. Salis has received the DARPA Young Faculty award and the NSF CAREER award for his achievements. He is also the founder of De Novo DNA, which runs a web-based design platform for engineering organisms, used by over 10000 researchers to design over 900000 genetic systems for diverse biotech applications. Prof. Salis earned his B.S. in Chemical Engineering from Rutgers University (2002) and his Ph.D. in Chemical Engineering from the University of Minnesota (2007). He was a postdoc at UCSF with Chris Voigt (2007-2009). He joined Penn State University in 2010.

Zoom:
https://umich.zoom.us/j/91712262512