Patrick Wilson of the University of Chicago in the U.S. will generate a universal influenza vaccine based on multiple conserved and protective viral protein epitopes selected for their ability to produce a broad and lasting immune response. Influenza causes tens of thousands of deaths worldwide each year, and current vaccines designed around individual epitopes are only 20% to 60% effective. To develop a more effective vaccine, they will use a large panel of human antibodies that bind the influenza virus to screen an assortment of hemagglutinin (HA) and neuraminidase (NA) viral proteins from different strains to identify conserved fragments that are the most immunogenic. These data will be combined with computational models to incorporate additional features such as epitope stability, and used to assemble a mosaic of HA and NA antigens from different influenza strains, which should elicit a more effective immune response, including long-term protection against multiple strains. The mosaic antigens will be iteratively tested during development, both in vitro and in vivo, in anticipation of pre-clinical testing and transition to clinical trials.