Influenza virus evolves under selective pressure from immunity across the human population, resulting in a progressive variation of viral antigenicity and therefore limited vaccine efficacy.  Our goal is to establish a secure mechanistic foundation for novel vaccination strategies that might confer long-lasting immunity in face of rapid antigenic variation of currently circulating subtypes and in anticipation of potential introduction of zoonotic subtypes. In other words, if we know the kind of antibody we would like to elicit, how do we learn enough about the immune response to create the right immunogen and the right immunization regimen?