skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

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

Abstract

Understanding the binding between human leukocyte antigens (HLAs) and peptides is important to understand the functioning of the immune system. Since it is time-consuming and costly to measure the binding between large numbers of HLAs and peptides, computational methods including machine learning models and network approaches have been developed to predict HLA-peptide binding. However, there are several limitations for the existing methods. We developed a network-based algorithm called sNebula to address these limitations. We curated qualitative Class I HLA-peptide binding data and demonstrated the prediction performance of sNebula on this dataset using leave-one-out cross-validation and five-fold cross-validations. Furthermore, this algorithm 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. We believe sNebula is an effective method to predict HLA-peptide binding and thus improve our understanding of the immune system.

Authors:
 [1];  [2];  [2];  [2];  [2];  [2]
  1. U.S. Food and Drug Administration, Jefferson, AR (United States); Univ. of Arkansas at Little Rock/Univ. of Arkansas for Medical Sciences Bioinformatics Graduate Program, Little Rock, AR (United States)
  2. U.S. Food and Drug Administration, Jefferson, AR (United States)
Publication Date:
Research Org.:
Oak Ridge Inst. for Science and Education (ORISE), Oak Ridge, TN (United States); U.S. Food and Drug Administration, Jefferson, AR (United States). National Center for Toxicological Research
Sponsoring Org.:
USDOE
OSTI Identifier:
1378368
Resource Type:
Journal Article: Accepted Manuscript
Journal Name:
Scientific Reports
Additional Journal Information:
Journal Volume: 6; Journal Issue: 1; Journal ID: ISSN 2045-2322
Publisher:
Nature Publishing Group
Country of Publication:
United States
Language:
English
Subject:
97 MATHEMATICS AND COMPUTING; 60 APPLIED LIFE SCIENCES; MHC; Network topology

Citation Formats

Luo, Heng, Ye, Hao, Ng, Hui Wen, Sakkiah, Sugunadevi, Mendrick, Donna L., and Hong, Huixiao. sNebula, a network-based algorithm to predict binding between human leukocyte antigens and peptides. United States: N. p., 2016. Web. doi:10.1038/srep32115.
Luo, Heng, Ye, Hao, Ng, Hui Wen, Sakkiah, Sugunadevi, Mendrick, Donna L., & Hong, Huixiao. sNebula, a network-based algorithm to predict binding between human leukocyte antigens and peptides. United States. doi:10.1038/srep32115.
Luo, Heng, Ye, Hao, Ng, Hui Wen, Sakkiah, Sugunadevi, Mendrick, Donna L., and Hong, Huixiao. 2016. "sNebula, a network-based algorithm to predict binding between human leukocyte antigens and peptides". United States. doi:10.1038/srep32115. https://www.osti.gov/servlets/purl/1378368.
@article{osti_1378368,
title = {sNebula, a network-based algorithm to predict binding between human leukocyte antigens and peptides},
author = {Luo, Heng and Ye, Hao and Ng, Hui Wen and Sakkiah, Sugunadevi and Mendrick, Donna L. and Hong, Huixiao},
abstractNote = {Understanding the binding between human leukocyte antigens (HLAs) and peptides is important to understand the functioning of the immune system. Since it is time-consuming and costly to measure the binding between large numbers of HLAs and peptides, computational methods including machine learning models and network approaches have been developed to predict HLA-peptide binding. However, there are several limitations for the existing methods. We developed a network-based algorithm called sNebula to address these limitations. We curated qualitative Class I HLA-peptide binding data and demonstrated the prediction performance of sNebula on this dataset using leave-one-out cross-validation and five-fold cross-validations. Furthermore, this algorithm 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. We believe sNebula is an effective method to predict HLA-peptide binding and thus improve our understanding of the immune system.},
doi = {10.1038/srep32115},
journal = {Scientific Reports},
number = 1,
volume = 6,
place = {United States},
year = 2016,
month = 8
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record

Save / Share:
  • We selected the estrogen receptor protein for study because of prior results indicating that arsenite is a 'potential nonsteroidal environmental estrogen'. We utilized radioactive {sup 73}As-labeled arsenite and vacuum filtration methodology to determine the binding affinity of arsenite to synthetic peptides. A zinc finger region containing four free sulfhydryls and the hormone binding region containing three free sulfhydryls based on the human estrogen receptor-{alpha} were studied. Peptide 15 (RYCAVCNDYASGYHYGVWSCEGCKA) bound arsenite with a K {sub d} of 2.2 {mu}M and B {sub max} (maximal binding capacity) of 89 nmol/mg protein. Peptide 10 (LECAWQGKCVEGTEHLYSMKCKNV) had a K {sub d} of 1.3more » {mu}M and B {sub max} of 59 nmol/mg protein. In contrast, the monothiol peptide 19 (LEGAWQGKGVEGTEHLYSMKCKNV) bound arsenite with a higher K {sub d} of 124 {mu}M and a B {sub max} of 26 nmol/mg protein. In our studies, amino acids other than cysteine (including methionine and histidine) did not bind arsenite at all. Peptides modeled on the estrogen receptor with two or more nearby free sulfhydryls (two or five intervening amino acids) had low K {sub d} values in the 1-4 {mu}M range. Peptides containing single sulfhydryls or two sulfhydryls spaced 17 amino acids apart had higher K {sub d} values in the 100-200 {mu}M range, demonstrating lower affinity. With the exception of peptide 24 which had an unusually high B {sub max} value of 234 nmol/mg, the binding capacity of the studied peptides was proportional to the number of free cysteines. Binding of trivalent arsenicals to peptides and proteins can contribute to arsenic toxicity and carcinogenicity via altered peptide/protein structure and enzyme function.« less
  • Human leukocyte antigen (HLA) gene polymorphism plays a critical role in protective immunity, disease susceptibility, autoimmunity, and drug hypersensitivity, yet the basis of how HLA polymorphism influences T cell receptor (TCR) recognition is unclear. We examined how a natural micropolymorphism in HLA-B44, an important and large HLA allelic family, affected antigen recognition. T cell-mediated immunity to an Epstein-Barr virus determinant (EENLLDFVRF) is enhanced when HLA-B*4405 was the presenting allotype compared with HLA-B*4402 or HLA-B*4403, each of which differ by just one amino acid. The micropolymorphism in these HLA-B44 allotypes altered the mode of binding and dynamics of the bound viralmore » epitope. The structure of the TCR-HLA-B*4405EENLLDFVRF complex revealed that peptide flexibility was a critical parameter in enabling preferential engagement with HLA-B*4405 in comparison to HLA-B*4402/03. Accordingly, major histocompatibility complex (MHC) polymorphism can alter the dynamics of the peptide-MHC landscape, resulting in fine-tuning of T cell responses between closely related allotypes.« less
  • The production and crystallization of human leukocyte antigen class II molecules HLA-DQ2 and HLA-DQ8 in complex with deamidated gliadin peptides is reported. Crystals of HLA-DQ2{sup PQPELPYPQ} diffracted to 3.9 Å, while the HLA-DQ8{sup EGSFQPSQE} crystals diffracted to 2.1 Å, allowing structure determination by molecular replacement. The major histocompatibility complex (MHC) class II molecules HLA-DQ2 and HLA-DQ8 are key risk factors in coeliac disease, as they bind deamidated gluten peptides that are subsequently recognized by CD4{sup +} T cells. Here, the production and crystallization of both HLA-DQ2 and HLA-DQ8 in complex with the deamidated gliadin peptides DQ2 α-I (PQPELPYPQ) and DQ8more » α-I (EGSFQPSQE), respectively, are reported.« less
  • Sister chromatid exchange (SCE) induction by ultraviolet (UV) light was studied in both human and pig whole blood cultures (WBC) and plasma leukocyte cultures (PLC). No variation in SCE frequency was observed between pig WBC and PLC in control as well as in treated cells. Conversely, SCE frequencies of human PLC were consistently higher than those of WBC in control and UV-exposed cells. Thus, red blood cells (RBCs) do not influence the sensitivity of lymphocytes to UV LIGHT exposure, and there must be some different culture condition(s) in the inducation of SCEs between human WBC and PLC but not inmore » swine lymphocyte cultures. Since the BrdUrd/lymphocyte ratio of WBC was halved in PLC, the effect of BrdUrd concentration in inducing the SCE baseline frequency of PLC may be ruled out. Neither the cell separation technique nor polymorphonuclear leukocytes had a significant role in the elevated SCE frequency of human PLC or MLC. Experiments where human RBCs were titrated into human PLC showed that the induction of an elevated SCE frequency of PLC was suppressed in a dose-dependent manner by the presence of RBCs in the culture medium. Since the incorporation of pig or human RBCs into human PLC as well as into MLC reduced the SCE frequency to that of WBC, a common component and/or function existing in these cells is suggested. Analysis of different RBC components showed that RBCs, specifically RBC ghosts, release a diffusible but not dialyzable corrective factor into culture medium that is able to reduce the SCE frequencies of PLC.« less