GraphMHC: neoantigen prediction model applying the graph neural network to molecular structure

Abstract

Neoantigens are biomarkers that can predict the prognosis associated with immune checkpoint inhibition by estimating the binding potential of candidate peptides to somatic mutation and major histocompatibility complex (MHC) proteins. Although deep neural networks have been primarily used for these prediction models, it is difficult to consider the models reported thus far as accurately representing the interactions between biomolecules. In this study, we propose the GraphMHC model, which utilizes a graph neural network model through molecular structure to simulate the binding between MHC proteins and peptide sequences. Amino acid sequences sourced from the immune epitope database (IEDB) undergo conversion into molecular structures. Subsequently, atomic intrinsic informations and inter-atomic connections are extracted and structured as a graph representation. Bindings are classified by feeding them into the GraphMHC network, comprising stacked graph attention and convolution layers. The prediction results from the test set using the GraphMHC model showed a high performance with an area under the receiver operating characteristic curve of 92.2% (91.9-92.5%), surpassing the baseline model. Moreover, by applying the GraphMHC model to melanoma patient data from the Cancer Genome Atlas project, we found a borderline difference in overall survival and a significant difference in stromal score between the high and low neoantigen load groups. This distinction was not present in the baseline model. This study presents the first feature-intrinsic method based on biochemical molecular structure for modeling the binding between MHC protein sequences and neoantigen candidate peptide sequences. The model can provide highly accurate suitability information for cancer patients who want to apply immune checkpoint inhibitors. Author summary