Andrew Ferguson receives award to design and engineer proteins using machine learning

Assoc. Prof. Andrew Ferguson has received a 2020 Machine Learning in the Chemical Sciences and Engineering Award, a new honor presented by the Camille and Henry Dreyfus Foundation that recognizes projects with potential to contribute new insight and innovation to the field.

The award supports a joint research project between Ferguson and Prof. Rama Ranganathan in data-driven protein engineering.

“It's a great honor to be recognized as part of the freshman class for this new awards program from the Dreyfus Foundation,” said Ferguson. “I was delighted to have been recognized by such a venerable foundation that has long stood at the forefront of funding innovative chemical science research.”

In their project, Ferguson and Ranganathan aim to both learn "nature's blueprint" for protein design using deep generative machine learning models and to discover proteins with novel or improved function using high-throughput gene synthesis and assays. The researchers have founded a startup, Evozyne, that will commercialize this technology with applications in energy, environment, catalysis, and agriculture.

Several iterations of their design-build-test cycle, in which experimental measurements are passed back to retrain the computational models within a virtuous feedback loop, will mimic millions of years of evolution in the laboratory.

“Natural selection has explored a vanishing fraction of possible proteins,” said Ferguson, “and designing proteins for synthetic applications amounts to divining needles in the vast haystack of sequences unexplored by nature.”

The rational design of synthetic proteins with tailored function is a long-standing goal in chemical science, with broad applications in diverse areas of chemical engineering, materials science, and public health. Ferguson and Ranganathan anticipate that the machine learning platform can help them design and engineer proteins with new functions and capabilities that do not exist in nature.

“For example, imagine being able to engineer biological enzymes to replace energy-intensive chemical reactions or separations, novel antibodies to combat infectious disease, or entirely new biomaterials to replace oil-based polymers,” said Ferguson.