scholarly journals A comprehensive review and performance evaluation of bioinformatics tools for HLA class I peptide-binding prediction

2020 ◽  
Vol 21 (4) ◽  
pp. 1119-1135 ◽  
Author(s):  
Shutao Mei ◽  
Fuyi Li ◽  
André Leier ◽  
Tatiana T Marquez-Lago ◽  
Kailin Giam ◽  
...  
Keyword(s):  
Machine Learning ◽  
T Cell ◽  
Peptide Binding ◽  
Hla Class I ◽  
Machine Learning Algorithms ◽  
Class I ◽  
Target Cells ◽  
Binding Prediction ◽  
Validation Data ◽  
Data Set

Abstract Human leukocyte antigen class I (HLA-I) molecules are encoded by major histocompatibility complex (MHC) class I loci in humans. The binding and interaction between HLA-I molecules and intracellular peptides derived from a variety of proteolytic mechanisms play a crucial role in subsequent T-cell recognition of target cells and the specificity of the immune response. In this context, tools that predict the likelihood for a peptide to bind to specific HLA class I allotypes are important for selecting the most promising antigenic targets for immunotherapy. In this article, we comprehensively review a variety of currently available tools for predicting the binding of peptides to a selection of HLA-I allomorphs. Specifically, we compare their calculation methods for the prediction score, employed algorithms, evaluation strategies and software functionalities. In addition, we have evaluated the prediction performance of the reviewed tools based on an independent validation data set, containing 21 101 experimentally verified ligands across 19 HLA-I allotypes. The benchmarking results show that MixMHCpred 2.0.1 achieves the best performance for predicting peptides binding to most of the HLA-I allomorphs studied, while NetMHCpan 4.0 and NetMHCcons 1.1 outperform the other machine learning-based and consensus-based tools, respectively. Importantly, it should be noted that a peptide predicted with a higher binding score for a specific HLA allotype does not necessarily imply it will be immunogenic. That said, peptide-binding predictors are still very useful in that they can help to significantly reduce the large number of epitope candidates that need to be experimentally verified. Several other factors, including susceptibility to proteasome cleavage, peptide transport into the endoplasmic reticulum and T-cell receptor repertoire, also contribute to the immunogenicity of peptide antigens, and some of them can be considered by some predictors. Therefore, integrating features derived from these additional factors together with HLA-binding properties by using machine-learning algorithms may increase the prediction accuracy of immunogenic peptides. As such, we anticipate that this review and benchmarking survey will assist researchers in selecting appropriate prediction tools that best suit their purposes and provide useful guidelines for the development of improved antigen predictors in the future.

scholarly journals HLA class I binding prediction via convolutional neural networks

2017 ◽  
Author(s):  
Yeeleng S. Vang ◽  
Xiaohui Xie
Keyword(s):  
Machine Learning ◽  
Network Architecture ◽  
Peptide Binding ◽  
Hla Class I ◽  
Class I ◽  
Machine Learning Techniques ◽  
Binding Potential ◽  
Distributed Representation ◽  
Binding Prediction ◽  
Peptide Binding Prediction

AbstractMany biological processes are governed by protein-ligand interactions. One such example is the recognition of self and non-self cells by the immune system. This immune response process is regulated by the major histocompatibility complex (MHC) protein which is encoded by the human leukocyte antigen (HLA) complex. Understanding the binding potential between MHC and peptides can lead to the design of more potent, peptide-based vaccines and immunotherapies for infectious autoimmune diseases.We apply machine learning techniques from the natural language processing (NLP) domain to address the task of MHC-peptide binding prediction. More specifically, we introduce a new distributed representation of amino acids, name HLA-Vec, that can be used for a variety of downstream proteomic machine learning tasks. We then propose a deep convolutional neural network architecture, name HLA-CNN, for the task of HLA class I-peptide binding prediction. Experimental results show combining the new distributed representation with our HLA-CNN architecture acheives 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. Codes are available at https://github.com/uci-cbcl/HLA-bind.

Improving T Cell Functional Assays Using Mono-specific HLA Class I/Peptide Coated APCs/Target Cells

2005 ◽  
Vol 28 (6) ◽  
pp. 634
Author(s):  
Philip Savage ◽  
Linda Barber ◽  
Sue Jordan
Keyword(s):  
T Cell ◽  
Hla Class I ◽  
Class I ◽  
Target Cells ◽  
Functional Assays

Leukemic Blasts Acting as Host Antigen Presenting Cells Trigger a Combined CD4 and CD8 Allo-Immune Response Directed Against Mismatched HLA Class I.

Blood ◽  
2007 ◽  
Vol 110 (11) ◽  
pp. 5030-5030
Author(s):  
Avital Amir ◽  
Renate S. Hagendoorn ◽  
Erik W.A. Marijt ◽  
Roelof Willemze ◽  
J.H. Frederik Falkenburg ◽  
...  
Keyword(s):  
Immune Response ◽  
T Cells ◽  
T Cell ◽  
Hla Class I ◽  
Cd8 T Cell ◽  
Class I ◽  
Mixed Chimerism ◽  
Target Cells ◽  
Hla Dr ◽  
Donor T Cells

Abstract Single HLA locus mismatched stem cell transplantation (SCT) is applied in patients with hematological malignancies who may benefit from allogeneic transplantation but lack an HLA-matched donor. Although HLA disparity between patient and donor increases the risk of developing GVHD, the relative risk of GVHD after single HLA locus mismatched SCT is only 1.5 fold. In view of the high frequency of allo-HLA reactive T-cells, which is about 1000-fold higher than the frequencies of minor histocompatibility antigen specific T-cells, this risk increase is lower than could be expected. Since almost all nucleated cells express HLA class I, one would expect all single HLA class I mismatched transplanted patients to develop severe GVHD. We hypothesized therefore that the presentation of the HLA class I mismatched allele on nucleated cells of the patient is not sufficient to elicit an effective allo-immune response. We characterized the allo-immune response in a patient with acute myeloid leukemia (AML) who was treated with a T-cell depleted SCT from a sibling donor who was HLA identical except for an HLA-A2 crossover. Six months after SCT, donor lymphocyte infusion (DLI) of 2.5*10e6 T-cells/kg was given for mixed chimerism comprising 99% T-cells of patient origin. No clinical response and no GVHD developed. Twelve months after SCT 95% of T-cells were still of patient origin, and AML relapse occurred with 9% blasts in bone marrow for which a second DLI containing 7.5*10e6 T-cells/kg was given. Five weeks after the DLI the patient died of grade IV GVHD. During the GVHD, conversion to donor chimerism developed. In peripheral blood of the patient 90% of CD8 and 40% of CD4 donor T-cells were activated as determined by HLA-DR expression. To analyze the nature of the immune response, the activated CD8 and CD4 donor T-cells were single cell sorted, expanded and tested for alloreactivity and HLA restriction using cytotoxicity and cytokine production assays against a panel of target cells blocked with different HLA-mAbs. 82% of the CD8 T-cell clones were alloreactive and restricted to the allo-HLA-A2. The response was highly polyclonal as shown by the usage of different T-cell receptor Vβ chains with different CDR3 sequences. 26% of the CD4 clones were alloreactive and this response was also polyclonal. The CD4 clones were HLA-DR1 restricted and recognized donor EBV-LCL transduced with HLA-A2, indicating that the peptide recognized in HLA-DR1 was derived from the mismatched HLA-A2 molecule. The recognized epitope was demonstrated to comprise AA 103–120 derived from a hypervariable region of HLA-A2. At the time of the first DLI, only HLA class I expressing T-cells and non-hematopoietic patient derived cells were present, capable of activating the CD8 T-cells but not of triggering the CD4 response. Leukemic blasts present at the time of the second DLI, however, expressed both HLA-DR and HLA class I, and were shown to activate the CD4 as well as the CD8 clones. We hypothesize that the HLA class II expression on hematopoietic cells of the patient at the time of the relapse was essential for the development of this immune response. In conclusion, these results indicate a role for patient leukemic blasts acting as host APCs in initiating the GVH response by activating both a CD4 and CD8 T-cell response in an HLA class I mismatched setting.

scholarly journals The pockets guide to HLA class I molecules

2021 ◽  
Author(s):  
Andrea T. Nguyen ◽  
Christopher Szeto ◽  
Stephanie Gras
Keyword(s):  
T Cells ◽  
Cell Surface ◽  
Peptide Binding ◽  
Hla Class I ◽  
Class I ◽  
Side Chains ◽  
Binding Prediction ◽  
Peptide Residue ◽  
Hla Class I Molecules ◽  
Binding Pockets

Human leukocyte antigens (HLA) are cell-surface proteins that present peptides to T cells. These peptides are bound within the peptide binding cleft of HLA, and together as a complex, are recognised by T cells using their specialised T cell receptors. Within the cleft, the peptide residue side chains bind into distinct pockets. These pockets ultimately determine the specificity of peptide binding. As HLAs are the most polymorphic molecules in humans, amino acid variants in each binding pocket influences the peptide repertoire that can be presented on the cell surface. Here, we review each of the 6 HLA binding pockets of HLA class I (HLA-I) molecules. The binding specificity of pockets B and F are strong determinants of peptide binding and have been used to classify HLA into supertypes, a useful tool to predict peptide binding to a given HLA. Over the years, peptide binding prediction has also become more reliable by using binding affinity and mass spectrometry data. Crystal structures of peptide-bound HLA molecules provide a means to interrogate the interactions between binding pockets and peptide residue side chains. We find that most of the bound peptides from these structures conform to binding motifs determined from prediction software and examine outliers to learn how these HLAs are stabilised from a structural perspective.

A Novel Ensemble Stacking Classification of Genetic Variations Using Machine Learning Algorithms

2021 ◽  
Author(s):  
Jahnavi Yeturu ◽  
Poongothai Elango ◽  
S. P. Raja ◽  
P. Nagendra Kumar
Keyword(s):  
Machine Learning ◽  
Heart Diseases ◽  
Learning Algorithms ◽  
Genetic Mutation ◽  
Machine Learning Algorithms ◽  
Support Vector ◽  
Genetic Mutations ◽  
Validation Data ◽  
Data Set

Genetics is the clinical review of congenital mutation, where the principal advantage of analyzing genetic mutation of humans is the exploration, analysis, interpretation and description of the genetic transmitted and inherited effect of several diseases such as cancer, diabetes and heart diseases. Cancer is the most troublesome and disordered affliction as the proportion of cancer sufferers is growing massively. Identification and discrimination of the mutations that impart to the enlargement of tumor from the unbiased mutations is difficult, as majority tumors of cancer are able to exercise genetic mutations. The genetic mutations are systematized and categorized to sort the cancer by way of medical observations and considering clinical studies. At the present time, genetic mutations are being annotated and these interpretations are being accomplished either manually or using the existing primary algorithms. Evaluation and classification of each and every individual genetic mutation was basically predicated on evidence from documented content built on medical literature. Consequently, as a means to build genetic mutations, basically, depending on the clinical evidences persists a challenging task. There exist various algorithms such as one hot encoding technique is used to derive features from genes and their variations, TF-IDF is used to extract features from the clinical text data. In order to increase the accuracy of the classification, machine learning algorithms such as support vector machine, logistic regression, Naive Bayes, etc., are experimented. A stacking model classifier has been developed to increase the accuracy. The proposed stacking model classifier has obtained the log loss 0.8436 and 0.8572 for cross-validation data set and test data set, respectively. By the experimentation, it has been proved that the proposed stacking model classifier outperforms the existing algorithms in terms of log loss. Basically, minimum log loss refers to the efficient model. Here the log loss has been reduced to less than 1 by using the proposed stacking model classifier. The performance of these algorithms can be gauged on the basis of the various measures like multi-class log loss.

Ensemble approaches for improving HLA Class I-peptide binding prediction

2011 ◽  
Vol 374 (1-2) ◽  
pp. 47-52 ◽  
Author(s):  
Xihao Hu ◽  
Hiroshi Mamitsuka ◽  
Shanfeng Zhu
Keyword(s):  
Peptide Binding ◽  
Hla Class I ◽  
Class I ◽  
Binding Prediction ◽  
Peptide Binding Prediction

scholarly journals Regulation of T cell lymphokine production by killer cell inhibitory receptor recognition of self HLA class I alleles.

1996 ◽  
Vol 184 (2) ◽  
pp. 789-794 ◽  
Author(s):  
A D'Andrea ◽  
C Chang ◽  
J H Phillips ◽  
L L Lanier
Keyword(s):  
T Cells ◽  
T Cell ◽  
Mhc Class I ◽  
Cytokine Production ◽  
Killer Cell ◽  
Hla Class I ◽  
Class I ◽  
Target Cells ◽  
T Cell Clone ◽  
Mhc Class I Molecules

The killer cell inhibitory receptors (KIRs) are surface glycoproteins expressed by natural killer (NK) and T cells that specifically recognize defined groups of polymorphic human histocompatibility leukocyte antigen (HLA) class I molecules. Interactions between KIRs on NK or T cells and major histocompatibility complex (MHC) class I molecules on potential target cells inhibit cell-mediated cytotoxicity, presumably by delivering a negative signal preventing lymphocyte activation. In this study we examined whether KIRs also regulate cytokine production induced in response to T cell receptor-dependent T cell activation. CD4+ and CD8+ T cell clones were stimulated by bacterial superantigens in the presence or absence of monoclonal antibodies (mAbs) against the KIR NKB1 or MHC class I molecules, and production of tumor necrosis factor alpha and interferon gamma was evaluated. When bacterial superantigen was presented by an autologous antigen-presenting cell (APC) to a KIR+ T cell clone, cytokine production was always enhanced in the presence of anti-MHC class I mAb. Similarly, anti-KIR mAb also augmented cytokine production, provided that the APC expressed a HLA class I allele recognized by the KIR. These results suggest that recognition of autologous MHC class I molecules by KIR+ T cells provides a regulatory mechanism acting to modulate the potency of their responses to antigenic challenge.

scholarly journals HLA-derived peptides which inhibit T cell function bind to members of the heat-shock protein 70 family.

1996 ◽  
Vol 183 (2) ◽  
pp. 339-348 ◽  
Author(s):  
E Nossner ◽  
J E Goldberg ◽  
C Naftzger ◽  
S C Lyu ◽  
C Clayberger ◽  
...  
Keyword(s):  
T Cell ◽  
Heat Shock ◽  
T Lymphocytes ◽  
Heat Shock Protein ◽  
Cell Function ◽  
Peptide Binding ◽  
Hla Class I ◽  
Class I ◽  
Inhibitory Effects ◽  
Hla Class I Molecules

Synthetic peptides corresponding to sequences of HLA class I molecules have inhibitory effects on T cell function. The peptides investigated in this study have sequences corresponding to the relatively conserved region of the alpha 1 helix of HLA class I molecules that overlaps the "public epitope" Bw4/Bw6. These HLA-derived peptides exhibit inhibitory effects on T lymphocytes and have beneficial effects on the survival of allogenic organ transplants in mice and rats. Peptides corresponding to the Bw4a epitope appear most potent as they inhibit the differentiation of T cell precursors into mature cytotoxic T lymphocytes (CTL) and target cell lysis by established CTL lines and clones. To elucidate the mechanism through which these peptides mediate their inhibitory effect on T lymphocytes, peptide binding proteins were isolated from T cell lysates. We show that the inhibitory Bw4a peptide binds two members of the heat-shock protein (HSP) 70 family, constitutively expressed HSC70 and heat-inducible HSP70. Peptide binding to HSC/HSP70 is sequence specific and follows the rules defined by the HSC70 binding motif. Most intriguing, however, is the strict correlation of peptide binding to HSC/HSP70 and the functional effects such that only inhibitory peptides bind to HSC70 and HSP70 whereas noninhibitory peptides do not bind. This correlation suggests that small molecular weight HLA-derived peptides may modulate T cell responses by directly interacting with HSPs. In contrast to numerous reports of HSP70 expression at the surface of antigen-presenting cells and some tumor cells, we find no evidence that HSC/HSP70 are expressed at the surface of the affected T cells. Therefore, we believe that the peptides' immunodulatory effects are not mediated through a signaling event initiated by interaction of peptide with surface HSP, but favor a model similar to the action of other immunomodulatory compounds, FK506 and cyclosporin A, with a role for HSC/HSP70 similar to that for immunophilins, FKBPs and CyP40.

Primary Allogeneic T-Cell Responses Against In Vitro Modified Mantle Cell Lymphoma (MCL) Cells: Implications for Adoptive Immunotherapy after HLA-Matched Stem Cell Transplantation (SCT).

Blood ◽  
2004 ◽  
Vol 104 (11) ◽  
pp. 2118-2118
Author(s):  
M. Hoogendoorn ◽  
J. Olde Wolbers ◽  
W. M. Smit ◽  
I. Jedema ◽  
M. R. Schaafsma ◽  
...  
Keyword(s):  
T Cell ◽  
T Cell Responses ◽  
Hla Class I ◽  
Class I ◽  
Target Cells ◽  
Cell Responses ◽  
Cell Clones ◽  
Cd40 Activation ◽  
Antigen Presenting

Abstract Allogeneic SCT is being explored as treatment modality for patients with advanced MCL. Complete sustained remissions have been observed after allogeneic SCT illustrating susceptibility of MCL cells to graft-versus-lymphoma (GVL) effect.To potentiate this GVL effect and to reduce graft-versus-host disease (GVHD) reactivity, adoptive transfer of in vitro-selected cytotoxic T cells (CTLs) with specificity for MCL or for hematopoiesis-restricted minor histocompatibility antigens could be an attractive approach. The lack of expression of costimulatory molecules on MCL cells hampers the generation of MCL-reactive T cell-responses. To transform MCL cells into efficient antigen-presenting cells (APCs) we tested the B-lineage specific activating cytokines (IL-4), the unique MCL proliferating cytokine (IL-10) and the ligand of toll like receptor 9, CpG.Furthermore, CD40 triggering using irradiated CD40-L transfected murine fibroblasts (tCD40L) in combination with the cytokines and CpG was examined. The expression of the costimulatory and adhesion molecules CD80, CD86, CD83, CD54 and CD58 of MCL cells of 7 patients, all carrying the t(11;14) translocation, was analyzed by flowcytometry. No upregulation of any of these molecules was observed using the cytokines or CpG. Ligation of CD40 on MCL cells caused a significant upregulation of CD54,CD58, CD80 and CD86 (p<0.01) with maximal expression after 4 days of stimulation. No additional upregulation was induced from IL- 4, IL -10 or CpG. The cumulative production of IL-12 and IL-10 by the MCL cells in response to the various stimuli after 4 days was measured. High amounts of IL-12 (median 1640 pg/mL, range 67–8800 pg/mL) in the absence of IL-10(<100 pg/mL) were synthesized by MCL cells after CD40 activation. Additional stimulation with CpG enhanced the production of IL-12 (1870 pg/mL, range 77–30000 pg/mL) but also the production of IL-10(299 pg/mL, range 0–418 pg/mL). MCL cells were unable to produce IL-12 without CD40 triggering (<5 pg/mL). To analyze the antigen-presenting capacity of primary MCL cells as well as CD40-activated MCL cells (MCL-APC), CD8+ T cells from an unrelated HLA-A and B matched and from a HLA-class I matched donor were stimulated with MCL or MCL-APC cells. Primary MCL cells were not capable of generating T-cell lines. Using a newly developed flowcytometry-based cytotoxicity assay in which the target cells were labeled with CFSE (Jedema,Blood2004;103:2677) we investigate whether the CTL lines, generated against MCL-APC were cytotoxic against MCL-specific targets. The CD8+ CTL lines from both donors effectively killed at an E/T ratio of 10:1 primary MCL (53%) and MCL-APC (83%) and not PHA blasts from the donor. Using limiting dilution assay, in both donor/patient pairs MCL-reactive CTL clones could be generated. 60 out of 89 proliferating CD8+ T cell clones from the first patient/donor pair and 29 out of 74 proliferating CD8+ T cell clones from the second combination showed specific lysis of primary MCL, MCL-APC and PHA blasts from the patient and not of PHA blasts of the donor. Blocking studies using anti-HLA class I antibodies of both CTL lines and clones confirmed class I restricted recognition of the target cells. In conclusion, CD40 activation transforms MCL cells into malignant APC, capable of producing high levels of IL-12 and capable of inducing vigorous MCL-reactive T-cell responses.

Sign in / Sign up

Export Citation Format

Share Document