Diabetes is a complex metabolic disease in which patients experience high blood glucose levels leading to a plethora of symptoms and long-term complications. In type 1 diabetes (T1D), this happens likely due to an autoimmune reaction targeting the insulin producing cell population, so-called beta cells, within the pancreatic islets. CD8+ T-cells are considered the main cell type implicated in beta cell destruction. However, their phenotype, recognized antigens, as well as how they interact with beta cells is still rather unknown. During disease progression, islets with a heightened expression of HLA class I (HLA-I) proteins can be observed. This feature is usually detected in all cells within one islet but not in all islets within one donor. The increased expression of HLA-I molecules on the surface of islet cells, and especially of beta cells, could be a precipitating factor and actively contribute to the presentation of self-antigens to the immune system. Based on this, we propose to investigate the differences at the proteomic level between islets with low, medium and high expression of HLA-I molecules on a cell-type resolved basis (in different cell types) using our recently published method Deep Visual Proteomics (DVP). This platform combines artificial-intelligence-driven image analysis of cellular phenotypes with automated single-cell or single-nucleus laser microdissection and ultrahigh-sensitivity mass spectrometry. Our goal is to analyze the proteomes of single-cell populations within their spatial context to provide answers regarding the significance of HLA-I upregulation with respect to disease etiology, as well as the proteomic profile of different islet cells during disease progression.