GRIP: Graph Representation of Immune Repertoire Using Graph Neural Network and Transformer

Abstract

The immune repertoire is a collection of immune receptors that has emerged as an important biomarker for both diagnostic and therapeutic of cancer patients. In terms of deep learning, analyzing immune repertoire is a challenging multiple-instance learning problem in which the immune repertoire of an individual is a bag, and the immune receptor is an instance. Although several deep learning methods for immune repertoire analysis are introduced, they consider the immune repertoire as a set-like structure that doesn’t take account of the nature of the immune response. When the immune response occurs, mutations are introduced to the immune receptor sequence sequentially to optimize the immune response against the pathogens that enter our body. As a result, immune receptors for the specific pathogen have the lineage of evolution; thus, immune repertoire is better represented as a graph-like structure. In this work, we present our novel method graph representation of immune repertoire (GRIP), which analyzes the immune repertoire as a hierarchical graph structure and utilize the collection of graph neural network followed by graph pooling and transformer to efficiently represents the immune repertoire as an embedding vector. We show that GRIP predict the survival probability of cancer patients better than the set-based methods and graph-based structure is critical for performance. Also, GRIP provides interpretable results, which prove that GRIP adequately use the prognosis-related immune receptor and give further possibility to use the GRIP as the novel biomarker searching tool.