In Silico: Predicting Intrinsic Features of HLA Class-I Restricted Neoantigens.

ABSTRACT

Mutation-containing immunogenic peptides from tumor cells, also named as neoantigens, have various amino acid descriptors and physical-chemical properties characterized intrinsic features, which are useful in prioritizing the immunogenicity potentials of neoantigens and predicting patients' survival. Here, we describe a glioma neoantigen intrinsic feature database, GNIFdb, that hosts computationally predicted HLA-I restricted neoantigens of gliomas, their intrinsic features, and the tools for calculating intrinsic features and predicting overall survival of gliomas. We illustrate the application of GNIFdb in searching for possible neoantigen candidates from ATF6 that plays important roles in tumor growth and resistance to radiotherapy in glioblastoma. We also demonstrate the application of intrinsic feature associated tools in GNIFdb to predict the overall survival of primary IDH wild-type glioblastoma.