Thomas Kepler wants to engineer vaccines just like others have engineered semiconductors.
Doing so requires understanding how vaccines work and why they sometimes fail. And doing that requires quantitative tools like the ones Kepler, a Duke University computational immunologist and former SFI Vice President, is building. In August he is leading a workshop at SFI on “Quantitative and Systems Biology” to introduce graduate students and postdocs to these new techniques.
One of the methods his team has developed provides a kind of X-ray view into the workings of immune cells, including dendritic cells. Dendritic cells are couriers that are activated by the vaccine and then transport it into the lymph node, where the protective immune reaction begins. Some candidate vaccines fail because too few dendritic cells are activated to play their role. Immunologists use a technique called “ ow cytometry” to count the dendritic cells in a lymph node and see if they’ve carried the vaccine in.
Flow cytometry produces overwhelming volumes of data: millions of points in spaces of as many as 20 dimensions. To make sense of it, researchers have had to pick out dozens of two-dimensional views and make judgments on what they see. Computers, however, can analyze all these points and dimensions at once, picking out patterns people would never catch.
“With these tools, when a vaccine doesn’t work, we don’t have to just throw it out,” Kepler says. “Sometimes we can see that if we changed it a little, maybe we could make it work better.”