HLA class I binding prediction via convolutional neural networks

@article{Vang2017HLACI,
  title={HLA class I binding prediction via convolutional neural networks},
  author={Yeeleng Scott Vang and Xiaohui Xie},
  journal={bioRxiv},
  year={2017}
}
Motivation : Many biological processes are governed by protein‐ligand interactions. [] Key Result Experimental results show combining the new distributed representation with our HLA‐CNN architecture achieves state‐of‐the‐art results in the majority of the latest two Immune Epitope Database (IEDB) weekly automated benchmark datasets. We further apply our model to predict binding on the human genome and identify 15 genes with potential for self binding. Availability and Implementation: Codes to generate the…
View on Wolters Kluwer
[PDF] Semantic Reader
81 Citations
Highly Influential Citations
6
Background Citations
9
Methods Citations
16
Results Citations
1

81 Citations

CcBHLA: pan-specific peptide–HLA class I binding prediction via Convolutional and BiLSTM features

This paper regards peptide fragments as natural language, and combines textCNN and BiLSTM to build a deep neural network model to encode the sequence features of HLA and peptides, demonstrating that the CcBHLA model outperforms the state-of-the-art known methods in detecting HLA class I binding peptides.

Prediction of peptide binding to MHC Class I proteins in the age of deep learning

The utility of deep learning for in silico prediction of peptide binding affinity to major histocompatibiliy complex Type I molecules (pMHC-I binding) is explored and six best-in-class methods are identified, all of which make highly correlated predictions.

Ranking-based Convolutional Neural Network Models for Peptide-MHC Binding Prediction

Two allele-specific Convolutional Neural Network (CNN)-based methods named ConvM and SpConvM are developed to tackle the binding prediction problem and form the problem as to optimize the rankings of peptide-MHC bindings via ranking-based learning objectives.

AI-MHC: an allele-integrated deep learning framework for improving Class I & Class II HLA-binding predictions

Allele-Integrated MHC (AI-MHC), a deep learning architecture with improved performance over the current state-of-the-art algorithms in human Class I and Class II MHC binding prediction, and is suitable for sequence analytics where a frame of interest needs to be learned in a longer, variable length sequence.

Ranking-Based Convolutional Neural Network Models for Peptide-MHC Class I Binding Prediction

Two allele-specific Convolutional Neural Network-based methods named ConvM and SpConvM are developed to tackle the binding prediction problem and form the problem as to optimize the rankings of peptide-MHC bindings via ranking-based learning objectives.

A computational framework for deep learning-based epitope prediction by using structure and sequence information

    Younghoon KimJ. Lee S. Yu
    Biology, Computer Science
    2019 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)
  • 2019
A new convolutional neural network (CNN)-based epitope prediction model is developed by incorporating new features including protein sequence vectors, chemical and structural characteristics of an epitope and MHC molecule, to achieve better prediction performance than the conventional methods when validated with a benchmark dataset from the Immune Epitope Database (IEDB).

DeepSeqPan, a novel deep convolutional neural network model for pan-specific class I HLA-peptide binding affinity prediction

Evaluation on public benchmark datasets shows that the proposed DeepSeqPan model without HLA structural information in training achieves state-of-the-art performance on a large number of HLA alleles with good generalization capability.

DeepNetBim: deep learning model for predicting HLA-epitope interactions based on network analysis by harnessing binding and immunogenicity information

It is observed that the interactive association constituted by human leukocyte antigen (HLA)-peptide pairs can be regarded as a network in which each HLA and peptide is taken as a node, and a network-based deep learning method called DeepNetBim was developed as a pan-specific epitope prediction tool.

IConMHC: a deep learning convolutional neural network model to predict peptide and MHC-I binding affinity

An in silico method with a deep convolutional neural network model, iConMHC, to predict peptide MHC binding affinity, which is a pan-allele model that is capable of making predictions for all the MHC alleles.

In silico Antibody-Peptide Epitope prediction for Personalized cancer therapy

An in-silico approach to identify patient-specific APEs by applying complex networks diagnostics on a novel multiplex data structure as input for a deep learning model that minimizes the required training time and the number of parameters.
...

56 References

sNebula, a network-based algorithm to predict binding between human leukocyte antigens and peptides

SNebula is a network-based algorithm that can predict not only peptides of different lengths and different types of HLAs, but also the peptides or HLAs that have no existing binding data and thus improve the understanding of the immune system.

Understanding and predicting binding between human leukocyte antigens (HLAs) and peptides by network analysis

It is demonstrated that network analysis coupled with Nebula is an efficient approach to understand and predict HLA-peptide binding interactions and thus, could further the understanding of ADRs.

NetMHCpan, a method for MHC class I binding prediction beyond humans

It is shown that the NetMHCpan-2.0 method can accurately predict binding to uncharacterized HLA molecules, including HLA-C and Hla-G, and is demonstrated to accurately predict peptide binding to chimpanzee and macaque MHC class I molecules.

NetMHCpan-3.0; improved prediction of binding to MHC class I molecules integrating information from multiple receptor and peptide length datasets

A neural network-based machine-learning algorithm leveraging information across multiple receptor specificities and ligand length scales is developed and demonstrated how this approach significantly improves the accuracy for prediction of peptide binding and identification of MHC ligands.

High-order neural networks and kernel methods for peptide-MHC binding prediction

The experiments show that the proposed shallow HONN outperform the popular pre-trained deep neural network on most tasks, which demonstrates the effectiveness of modelling high-order feature interactions for predicting major histocompatibility complex-peptide binding.

Gapped sequence alignment using artificial neural networks: application to the MHC class I system

It is shown that prediction methods based on alignments that include insertions and deletions have significantly higher performance than methods trained on peptides of single lengths and that the NetMHC-4.0 method can learn the length profile of different MHC molecules.

Scrutinizing MHC-I Binding Peptides and Their Limits of Variation

This study demonstrates that sophisticated machine-learning algorithms excel at extracting fine-grained patterns from peptide sequence data and predicting MHC-I binding peptides, thereby considerably extending existing linear prediction models and providing a fresh view on the computer-based molecular design of future synthetic vaccines.

Reliable prediction of T‐cell epitopes using neural networks with novel sequence representations

The difference in predictive performance between the neural network methods and that of the matrix‐driven methods is found to be most significant for peptides that bind strongly to the HLA molecule, confirming that the signal of higher order sequence correlation is most strongly present in high‐binding peptides.

Automated benchmarking of peptide-MHC class I binding predictions

A framework for the automated benchmarking of peptide-MHC class I binding prediction tools is described, running weekly benchmarks on data that are newly entered into the Immune Epitope Database (IEDB), giving the public access to frequent, up-to-date performance evaluations of all participating tools.

Quantitative prediction of class I MHC/epitope binding affinity using QSAR modeling derived from amino acid structural information.

The quantitative models established by STR-MLR could be used to guide virtual combinational design and high throughout screening of CTL epitope and have many advantages, such as definite physiochemical indication, easier calculation and explanation, and good performances.
...