Machine learning optimization of peptides for presentation by class II MHCs

Abstract

T cells play a critical role in normal immune responses to pathogens and cancer and can be targeted to MHC-presented antigens via interventions such as peptide vaccines. Here, we present a machine learning method to optimize the presentation of peptides by class II MHCs by modifying the peptide's anchor residues. Our method first learns a model of peptide affinity for a class II MHC using an ensemble of deep residual networks, and then uses the model to propose anchor residue changes to improve peptide affinity. We use a high throughput yeast display assay to show that anchor residue optimization successfully improved peptide binding.

A high-throughput yeast display approach to profile pathogen proteomes for MHC-II binding

“…Library selections were consistent with previous peptide-MHC-II yeast display dissociation studies (Dai et al, 2021;Rappazzo et al, 2020). Yeast were washed into pH 7.2 PBS to a concentration with 1 µM 3C protease and incubated at room temperature for 45 minutes.…”

“…The previously described null library (Dai et al, 2021) was generated with a peptide encoded as “NNNTAANNNNNNNNNTAGNNNNNNNNNNNNTGANNNNNN”, where “N” indicates any nucleotide and encodes ten random amino acids and three stop codons. This library was similarly generated in yeast using electrocompetent RJY100 yeast.…”

“…Generating reliable and rapid data on peptide-MHC binding could be useful for identifying important clinical targets, including for optimized T cell epitopes in vaccine design (Dai et al, 2021;Keskin et al, 2019;Liu et al, 2020Liu et al, , 2021aMoise et al, 2015;Ott et al, 2017;Patronov and Doytchinova, 2013;Rosati et al, 2021). In fact, therapeutics to generate antigenspecific T cell responses have shown great promise in cancer (Keskin et al, 2019;Ott et al, 2017) and infectious disease (Gambino et al, 2021).…”