Our parent R01 is focused on investigating the role of local unfolding in mediating the biological function of proteins, and a significant goal over the past 20 years of this grant has been the use of experimental data to develop and refine our evolving model of protein structural fluctuations so as to investigate the evolution of new function and disease. The goal of this project supplement is to capitalize on a recent discovery directly stemming from this project, whereby we are able to identify the role of “antigen mimicry” in the SARS COV-2 (the causative agent of COVID-19), which results in autoimmunity to specific proteins in a subset of infected patients. Our unique approach is (to our knowledge) the only predictive model that allows us to identify thermodynamic similarity between structurally and chemically different protein sequences. We show that by generating a thermodynamic fingerprint for each protein product of the SARs-COV-2 virus, we can compare the fingerprint with the fingerprints of the entire human proteome and identify statistically significant matches. Our original hypothesis was that similar thermodynamic fingerprints will be recognized by a common polyclonal antibody response. To challenge this hypothesis, we identified a number of high-identity matches for numerous proteins of the SARS-COV-2, one such example protein is orf10, which is predicted to share a signature with the human protein CD53. To directly test this initial prediction, we utilized a commercially available “proteome on a chip” technology to screen for the ability of orf10-specific polyclonal antibodies to bind with every expressed human protein. Remarkably, polyclonal antibodies to orf10 cross-reacted specifically with CD53, thus validating our hypothesis. As one possible disease etiology of “long-COVID” involves the high instances for viral- induced autoimmunity, our approach is uniquely suited to address this issue mechanistically. Our approach not only identifies which human proteins are similar with each viral protein and thus which are potential candidates for auto-immunity, it also identifies the sequence elements most responsible for the high similarity. This capability not only provides the medical community with a starting point to target mechanistic studies, it allows us and others to investigate the effects of genotypic variation with the human population. Here we will; 1) use our approach to identify all predicted matches between SARS-COV-2 proteins and the human proteome, and 2) experimentally test these predictions using commercially available “proteome on a chip” technology. All of the computed matches and the experimental validation data will be made available on our well-established web-server.