High-Throughput Prediction of MHC Class I and II Neoantigens with MHCnuggets

Abstract

Computational prediction of binding between neoantigen peptides and major histocompatibility complex (MHC) proteins can be used to predict patient response to cancer immunotherapy. Current neoantigen predictors focus on in silico estimation of MHC binding affinity and are limited by low predictive value for actual peptide presentation, inadequate support for rare MHC alleles, and poor scalability to high-throughput data sets. To address these limitations, we developed MHCnuggets, a deep neural network method that predicts peptide-MHC binding. MHCnuggets can predict binding for common or rare alleles of MHC class I or II with a single neural network architecture. Using a long short-term memory network (LSTM), MHCnuggets accepts peptides of variable length and is faster than other methods. When compared with methods that integrate binding affinity and MHC-bound peptide (HLAp) data from mass spectrometry, MHCnuggets yields a 4-fold increase in positive predictive value on independent HLAp data. We applied MHCnuggets to 26 cancer types in The Cancer Genome Atlas, processing 26.3 million allele-peptide comparisons in under 2.3 hours, yielding 101,326 unique predicted immunogenic missense mutations (IMM). Predicted IMM hotspots occurred in 38 genes, including 24 driver genes. Predicted IMM load was significantly associated with increased immune cell infiltration (P < 2 Â 10 À16), including CD8 þ T cells. Only 0.16% of predicted IMMs were observed in more than 2 patients, with 61.7% of these derived from driver mutations. Thus, we describe a method for neoantigen prediction and its performance characteristics and demonstrate its utility in data sets representing multiple human cancers.

Deep Learning in Proteomics

“…[141] During the past few years, a number of deep learningbased methods have been developed that outperform traditional machine learning methods, including shallow neural networks, for peptide-MHC binding prediction ( Table 4). Among these algorithms, 14 (ConvMHC, [142] HLA-CNN, [143] DeepMHC, [144] DeepSeqPan, [145] MHCSeqNet, [146] MHCflurry, [147] DeepHLApan, [148] ACME, [149] EDGE, [137] CNN-NF, [150] DeepNeo, [151] DeepLigand, [152] MHCherryPan, [153] and DeepAttentionPan [141] ) are specific for MHC class I binding prediction, three (DeepSeqPanII, [154] MARIA, [138] and NeonMHC2 [139] ) are specific for MHC class II binding prediction, and four (AI-MHC, [155] MHCnuggets, [156] PUFFIN, [157] and USMPep [158] ) can make predictions for both classes. All four types of peptide encoding approaches illustrated in Figure 1 are used in these tools, with one-hot encoding and BLOSUM matrix encoding being the most frequently used methods (Table 4).…”

Prioritization of SARS-CoV-2 epitopes using a pan-HLA and global population inference approach

“…Peptide binding predictions were performed using the pVACbind tool from pVACtools and executed using the griffithlab/pvactools:1.5.7 (https://hub.docker.com/r/griffithlab/pvactools/) Docker image. Class I prediction algorithms included MHCflurry (v1.6.0) (O'Donnell et al, 2018), MHCnuggets (v2.3) (Shao et al, 2020), NetMHC (v4.0) (Andreatta and Nielsen, 2016), PickPocket (v1.1) (Zhang et al, 2009), SMM (v1.0) (Peters and Sette, 2005), and SMMPMBEC (v1.0) (Kim et al, 2009). Class II prediction algorithms included NetMHCIIpan (v4.0) (Reynisson et al, 2020), SMMalign (v1.1) (Nielsen et al, 2007), NNalign (v2.3) (Nielsen and Andreatta, 2017), and MHCnuggets (v2.3) (Shao et al, 2020).…”

“…The neoCOVID Explorer application was developed using the Shiny R package [REF] and deployed using RStudio-Connect. Table REAGENT MHCnuggets (v2.3) (Shao et al, 2020) NetMHC (v4.0) (Andreatta and Nielsen, 2016) PickPocket (v1.1) (Zhang et al, 2009) SMM (v1.0) (Peters and Sette, 2005) SMMPMBEC (v1.0) (Kim et al, 2009) NetMHCIIpan (v4.0) (Reynisson et al, 2020) SMMalign (v1.1) (Nielsen et al, 2007)…”

Putting into Perspective the Future of Cancer Vaccines: Targeted Immunotherapy

“…The immune-inflamed phenotype correlates generally with higher response rates to anti-PD-L1/PD-1 therapy 51,62,67,[69][70][71] , which suggests that biomarkers could be used as predictive tools. Most attention has been paid to PD-L1, which is thought to reflect the activity of effector T cells because it can be adaptively expressed by most cell types following exposure to IFN-γ 6, 82 .…”

“…Computational models have been developed to achieve this goal. [50][51][52][53] The third step is to select the format of delivery of the vaccine. Commonly used formats include long peptides and RNA.…”