Our research is aimed at understanding how natural proteins perform their biological functions, and how we can apply this information to engineer novel proteins. We use high-throughput experimentation and computational modeling to systematically dissect the molecular basis of protein function. This work is highly interdisciplinary and combines approaches from biochemistry, molecular biology, applied physics, engineering, and computer science.

Machine learning-driven protein engineering

Understanding how protein sequence maps to biochemical function is crucial for rationally engineering proteins with new and useful properties. This mapping involves an extraordinarily complex balance of numerous physical interactions, many of which are still not well understood. Recent advances in high-throughput experimentation have generated an explosion of structural, genomic, and functional protein data. This vast quantity of data is a valuable resource for understanding the molecular basis of protein function. We are developing machine learning-based tools to infer the relationship between protein sequence and function from experimental data. This top-down modeling approach allows the discovery of new biochemical mechanisms and provides exceptional predictive accuracy for protein design.

Microfluidic technologies for high-throughput biochemistry

We are developing high-throughput technologies for performing biological assays in water-in-oil microemulsions. Each picoliter-sized droplet can compartmentalize reactions containing different cells, biomolecules, or combinations of chemicals. Various microfluidic modules enable droplets to be loaded with single cells, incubated over a range of times and temperatures, injected with additional reagents, and sorted based on their optical properties. These droplet operations can be performed with full automation at kilohertz frequencies (1000/sec), providing a general platform for high-throughput biochemistry. This technology can be applied to the directed evolution of enzymes, functional metagenomics, DNA synthesis, and mapping of protein sequence-function relationships.