Specificity in antigen recognition is a cornerstone of the adaptive immune system. A critical component of this is the recognition by T cell receptors of peptide antigen presented on major histocompatibilty complex (MHC) proteins. T cells are capable of recognizing viral or tumor-associated peptide-MHC and distinguishing disease-relevant antigens from self epitopes with fine specificity. As a consequence of the extreme diversity of peptide-MHC antigens, T cell receptors are paradoxically capable of cross-reacting to chemically and structurally distinct peptide-MHC.Research into the biophysical and structural underpinnings of T cell receptor antigen specificity and cross-reactivity have revealed some correlates and dependence on structural features, however our understanding remains too limited to develop immunotherapeutics intended to recruit T cells to a tumor in a patient-specific manner. Here, we demonstrate and critically evaluate computational methods for characterizing peptide-MHC structure for the purpose of determining "difference from self," in a high-throughput manner toward the goal of improving target prioritization in therapeutic pipelines.