The new coronavirus outbreak that begin in December 2019 has created a global public health emergency. This has led to an intense search to identify factors that contribute to the susceptibility and severity of illness. We recently developed an array to identify antibody-binding epitopes for rhinoviruses. Data from these arrays can be combined with information about viral protein structure to identify highly immunogenic regions for respiratory viruses. We propose to expand this array to include linear epitopes that represent the entire proteome of SARSCoV-2 and all other common coronaviruses that infect humans (OC43, NL63, etc.). The study population will include children from the COAST, WISC and URECA birth cohort studies who are also participating in the HEROS SARS-CoV-2 surveillance study. As part of routine cohort activities, these children undergo serial sampling of blood and nasal secretions that we can analyze using the array to determine individual patterns of antiviral antibody epitope recognition. We hypothesize that the pattern and quantity of antibody specific for epitopes of common coronaviruses contributes to the susceptibility to SARS-CoV-2 infection and illness. We propose three specific aims that will utilize sera obtained from children before and after HEROS-confirmed infection with SARS-CoV-2. First, in specimens obtained pre-infection we will use the array to identify patterns of antibody epitope recognition to common childhood coronaviruses, assess cross-reactivity with SARS-CoV-2, and determine whether cross-reactivity is associated with protection against infection or illness. In the second aim, we will determine whether the diversity of antibody responses to common respiratory viruses is associated with a reduced risk of infection or illness. Finally, in the third aim we will describe antibody binding patterns before and after known COVID-19 cases to identify candidate regions that are immunogenic and neutralizing. To accomplish this aim, we will perform micro-neutralization assays (available in the BSL3 laboratory of Dr. Kristen Bernard, UW Madison) on convalescent sera or nasal secretions from children who developed symptomatic infection. This information will be analyzed together with pre- and post-infection array data using machine learning approaches to identify neutralizing epitopes. Identifying patterns of serologic responses that are crossprotective could help to identify susceptible individuals in the population and direct the design of vaccines to current and future viruses.