Beyond Binding: Personalized Multi-omic Neoantigen Identification

Johns Hopkins University, Baltimore MD

Investigators

Abstract

PROJECT SUMMARY This project addresses technological challenges in personalized neoantigen immunotherapy for cancer treatment, focusing on developing a more precise and broadly applicable neoantigen prediction algorithm by integrating multiomic data. The research aims to innovate in algorithm and software development, enhancing the selection and validation of neoantigens for immunotherapy. Despite the potential of personalized neoantigen therapies in oncology, their clinical adoption is hindered by the time-consuming and costly process of neoantigen selection and validation. Current computational tools for neoantigen prediction have limited effectiveness due to several factors, including inadequate training on immunologically validated neoantigens, a bias toward European HLA alleles, and the absence of multiomic patient data in algorithm training. Our project leverages a multidisciplinary team and a rich dataset of matched tumor and normal specimens from diverse patient populations, focusing on underrepresented HLA alleles, and the highly sensitive Mutation- Associated Neoantigen Functional Expansion of Specific T Cells (MANAFEST) assay for T cell activation and expansion. We propose to generate the first public dataset of immunologically assessed neoantigens with patient-matched multiomic sequence data, enhancing the precision of neoantigen prediction through advanced algorithmic methods and the integration of comprehensive biological data. The research design includes the creation of a Multiomic Neoantigen Algorithm Training and Assessment Resource and the development of a novel deep learning-based neoantigen prediction algorithm that utilizes patient-specific multiomic data. This approach aims to improve the prediction and clinical utility of neoantigen- based therapies and to provide community resources through open data and open-source software, promoting wider application and innovation in the field. By addressing these challenges, the project expects to substantially enhance the precision and applicability of neoantigen prediction tools, paving the way for their effective use in clinical settings and improving outcomes in cancer immunotherapy.