3pHLA-score improves structure-based peptide-HLA binding affinity prediction

Abstract

Binding of peptides to Human Leukocyte Antigen (HLA) receptors is a prerequisite for triggering immune response. Estimating peptide-HLA (pHLA) binding is crucial for peptide vaccine target identification and epitope discovery pipelines. Computational methods for binding affinity prediction can accelerate these pipelines. Currently, most of those computational methods rely exclusively on sequence-based data, which leads to inherent limitations. Recent studies have shown that structure-based data can address some of these limitations. In this work we propose a novel machine learning (ML) structure-based protocol to predict binding affinity of peptides to HLA receptors. For that, we engineer the input features for ML models by decoupling energy contributions at different residue positions in peptides, which leads to our novel per-peptide-position protocol. Using Rosetta’s ref2015 scoring function as a baseline we use this protocol to develop 3pHLA-score. Our per-peptide-position protocol outperforms the standard training protocol and leads to an increase from 0.82 to 0.99 of the area under the precision-recall curve. 3pHLA-score outperforms widely used scoring functions (AutoDock4, Vina, Dope, Vinardo, FoldX, GradDock) in a structural virtual screening task. Overall, this work brings structure-based methods one step closer to epitope discovery pipelines and could help advance the development of cancer and viral vaccines.

PLANET: A Multi-objective Graph Neural Network Model for Protein–Ligand Binding Affinity Prediction

“…Combination of molecular docking and scoring function can achieve a balance between accuracy and efficiency, which is the common choice for jobs in reality . The accuracy of traditional scoring functions, such as X-Score and AutoDock Vina, is largely limited by the simple mathematical forms adopted by them. , Machine learning methods without predefined formulas can be more flexible to fit the training data and obtain higher accuracy than traditional scoring functions . Δ Vina RF 20 , a random-forest (RF)-based model, achieved the best metrics in almost all tests in the CASF-2016 benchmark .…”

SARS-Arena: Sequence and Structure-Guided Selection of Conserved Peptides from SARS-related Coronaviruses for Novel Vaccine Development

“…For this reason, in the fifth and final step, we offer the opportunity to rescore the pHLAs generated by APE-Gen using different scoring functions. We added well-known scoring functions - Vina, Vinardo, and AD4 scoring - as well as a new machine learning-based scoring function recently developed, called 3pHLA ( 43 ).…”

“…Binding energy of modeled peptide-HLA structures can be evaluated within Workflow 2 using four integrated scoring functions: AutoDock4 ( 40 ), Vina ( 41 ), Vinardo ( 42 ) and 3pHLA-score ( 43 ). AutoDock4 score is based on an empirical free energy forcefield and is a part of a widely used protein-ligand docking tool.…”

APE-Gen2.0: Expanding Rapid Class I Peptide–Major Histocompatibility Complex Modeling to Post-Translational Modifications and Noncanonical Peptide Geometries

“…As such, future work will emphasize on using additional force field parameters, in order to expand the OpenMM energy minimization step to other PTMs. Additionally, as there have already been examples in the literature that use pMHC modeled structures to learn binding affinity or immunogenicity labels, future work will emphasize modeling a larger data set of phosphorylated peptides and use it in downstream tasks. Given that the scoring function alone could discern effects of the presence/absence of phosphorylation on the binding affinity (Figure ), we hypothesize that further fine-tuning scoring functions on specific binding affinity labels of phosphorylated/non-phosphorylated peptides can further improve performance.…”

“…Most of the methods that predict the binding affinity of the peptide to the MHC-I, the crucial first step in eliciting an immune response, have long been based on analyzing peptide sequences, , due to the large amounts of binding affinity and mass spectroscopy data that are publicly available . Methods that determine the immunogenicity of a peptide solely based on its amino-acid sequence have also started emerging rapidly. , In contrast to the availability of sequence data and sequence-based methods, the number of available pMHC crystal structures in public databases is order of magnitudes lower. , However, there is extensive evidence that structural features stemming from the bound peptide are predictive of properties such as binding affinity, stability, and peptide immunogenicity. , Certain chemical modifications such as single point mutations , or post-translational modifications (PTMs) such as phosphorylation can cause severe structural alterations, thus, noticeable effects in T-cell recognition, with minimum effect on the peptide sequence . Moreover, there have been studies which employed modeled pMHC structures and subsequently extracted structural features that have shown to be predictive of the aforementioned properties, even exhibiting competitive performance in comparison to peptide sequence-based tools. ,, It follows that devising algorithms and methodologies that provide accurate geometries of pMHC models is crucial in immune response-related tasks.…”