Abstract Neoantigens are tumor-specific molecules arising from somatic alterations in cancer cells and have garnered significant interest due to their immunogenic potential. Consequently, numerous computational pipelines have been developed to identify these targets. However, systematic comparisons between neopeptide generation tools are lacking, and there is no consensus on how to handle different mutation types. To address this gap, we compared the neopeptide sequences generated by four widely used tools: the Mutated Peptide Generator (MPG) from the Cancer Epitope Database and Analysis Resource (CEDAR), the Personalized Variant Antigens by Cancer Sequencing (pVACseq), the Mutated Peptide eXtractor and Informer (MuPeXI), and the Neoantigen Prediction Pipeline (NeoPredPipe). We applied these tools to somatic mutations from the Catalogue of Somatic Mutations in Cancer (COSMIC) v102 and validated the results with experimentally validated neoantigens curated in the CEDAR database. In total, 25% of the COSMIC mutations were considered by at least one method to generate neopeptides. The methods showed considerable variability, as only 22% of the neopeptides were generated by all tools. Overall, 29% of the discrepancies in neopeptide generation were attributed to different criteria used to select mutations or transcripts for downstream analysis. The remaining discrepancies were caused by differences in the algorithms used to handle and modify reference sequences into mutated neopeptides. Experimentally validated neoepitopes from CEDAR comprised only 0. 005% of the total generated neopeptides. While most neoepitopes were accurately generated by the four methods, 10% were not consistently identified across the tools. These findings underscore the need for methodological standardization to ensure reliable and reproducible neoantigen discovery. To our knowledge, this is the first comprehensive evaluation of neoantigen pipelines focused specifically on neopeptide sequence generation. Citation Format: Ibel Carri, Angela Frentzen Worley, Ashmitaa Logandha Ramamoorthy Premlal, Gauri Renjith, Malachi Griffith, Jason Greenbaum, Alessandro Sette, Bjoern Peters, Zeynep Kosaloglu-Yalcin. A comparative study of neoantigen discovery pipelines uncovers discrepancies in the generation of mutated neopeptide sequences abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts) ; 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86 (8Suppl): Abstract nr LB161.
Carri et al. (2026) studied this question.