Derek Smith (Photo: James King-Holmes)

As a Texas Instruments researcher working on artificial neural networks and speech recognition in the early 1990s, SFI Science Board member Derek Smith (University of Cambridge) was applying science to real-world problems every day. But he wanted to dig deeper.

“It seemed to me that the work that I was doing on pattern recognition and speech recognition might be related to how our immune systems recognize different strains of pathogens,” Smith explains, “so I started looking around to see where I might do such work.”

An invitation from SFI External Professors Stephanie Forrest (Arizona State University) and Alan Perelson (LANL) to join Forrest’s Ph.D. program at the University of New Mexico came with a suggestion: First, attend the Complex Systems Summer School at SFI.

One month later, Smith took a leave of absence from industry and began the summer program. The experience was transformative.

“The first morning, Mike Simmons, VP for Academic Affairs, talked about how at SFI there were no barriers between disciplines, and I was completely and utterly gobsmacked,” Smith says. “I knew this was the place for me to explore my ideas.”

After his studies at SFI, Smith went on to work in industry at Popular Power to develop these ideas, worked closely with US public health colleagues at the US Centers for Disease Control, and eventually ended up as a full professor at Cambridge University.

As a direct result of Smith’s work at SFI, he and colleagues Ron Fouchier and Alan Lapedes developed a method to understand the evolution of viruses they called antigenic cartography, and their work was published in Science (2004).

“We looked at the evolution of a virus in a new way, in particular, how it escapes our immune response,” Smith explains. “Diseases like HIV, malaria, and influenza persist and are very difficult to create vaccines for because viruses can change their surface proteins, making them much more difficult to understand.”

Soon after the paper was published, Smith and team were invited to apply their work to the approximately 20,000 influenza strains analyzed each year by public health laboratories for the World Health Organization. Their objective? Learn how the virus was evolving to identify which strains of flu should go in the vaccine — a process he has been integrally involved in since that first invitation.

“We felt that it was our responsibility to do this work because people’s lives are on the line,” Smith says. “But we also recognized that by applying our work, there was a real opportunity. The possibility to see the complete global evolution of the virus, in real-time, is an evolutionary biologist’s dream.”

Empowered by this global data and $24 million in funding from NIH and BARDA, Smith and his colleagues are now working to understand this evolution well enough to predict it and apply it to the flu vaccine.

“Derek’s success with his vaccine work is a perfect example of the insights that can be found in the liminal space between the traditional domains of academia, industry, and public health,” says Will Tracy, SFI’s VP for Applied Complexity.