HLA class I binding prediction via convolutional neural networks

摘要

Many biological processes are governed by protein-ligand interactions. Onesuch example is the recognition of self and nonself cells by the immune system.This immune response process is regulated by the major histocompatibilitycomplex (MHC) protein which is encoded by the human leukocyte antigen (HLA)complex. Understanding the binding potential between MHC and peptides can leadto the design of more potent, peptide-based vaccines and immunotherapies forinfectious autoimmune diseases. We apply machine learning techniques from the natural language processing(NLP) domain to address the task of MHC-peptide binding prediction. Morespecifically, we introduce a new distributed representation of amino acids,name HLA-Vec, that can be used for a variety of downstream proteomic machinelearning tasks. We then propose a deep convolutional neural networkarchitecture, name HLA-CNN, for the task of HLA class I-peptide bindingprediction. Experimental results show combining the new distributedrepresentation with our HLA-CNN architecture achieves state-of-the-art resultsin the majority of the latest two Immune Epitope Database (IEDB) weeklyautomated benchmark datasets. We further apply our model to predict binding onthe human genome and identify 15 genes with potential for self binding.