Updated project metadata. HLA-DQ molecules can be formed as both cis and trans variants. So far, the progress for predicting HLA-DQ antigen presentation has been limited. In addition, the contribution of trans-only variants in shaping the HLA-DQ immunopeptidome remains largely unresolved. Here, we address these issues by integrating state-of-the-art immunoinformatics data mining models with large volumes of high-quality HLA-DQ specific immunopeptidomics data. The analysis demonstrated a highly improved predictive power and molecular coverage for models trained with these novel HLA-DQ data and a limited to no contribution for trans-only HLA-DQ variants to the overall HLA-DQ immunopeptidome.