@article{hess_lee_holmes_zaharoff_2025, title={Assessing mutation-origin neopeptide prediction tools in canine cancer using the model MHC class Ia allomorphs, DLA-88*508:01 and *034:01 9302}, DOI={10.1093/jimmun/vkaf283.2724}, abstractNote={Abstract Description Pet dogs with osteosarcoma and bladder cancer are valid models for immunotherapy development. Lacking are any methodologic studies for finding canine neoantigen CTL. Our hypothesis is that actionable, mutant neopeptides are contained in canine cancer immunopeptidomes. We used standard in silico analyses and pMHC surface stabilization (pMHC-SS) assays to validate tools to identify neopeptides and cognate CTL. Two class Ia (DLA-88) allomorphs, prevalent in cancer-prone breeds and used to train NetMHCpan, were studied. Binding scores of viral MHC-associated peptides (MAPs) compared to NetMHCpan outputs established the best correlated parameter (BA score) and predictive cut-off value. In tests of predictive power, most putative binding neopeptides, generated from a published mammary tumor mutanome, bound in the flow assay. NetCHOP provided added benefits, despite overlapping c-terminal preferences of the allomorphs and canine proteasome. Rigor of the optimized tools to predict mutant MAPs for a model canine sarcoma was assessed comparing exome, transcriptome, and mass spectrometry data. Twenty-two mutant MAPs were predicted; none were found. The pMHC-SS assay was also used to evaluate peptide exchange in DLA-88 for custom neoepitope tetramers. Motif-matching dipeptides and low-affinity leaving peptides facilitated swapping (100% efficiency). Conventional neopeptide prediction methods are helpful in dogs. The rarity of mutant MAPs is a challenge to immunotherapy in both species. Funding Sources NIH U01CA272258 Topic Categories Veterinary and Comparative Immunology (VET)}, journal={The Journal of Immunology}, author={Hess, Paul R. and Lee, Ching-Yen and Holmes, Jennifer and Zaharoff, David}, year={2025}, month={Nov} }