PMID- 30408041
OWN - NLM
STAT- MEDLINE
DCOM- 20190314
LR  - 20190314
IS  - 1553-7358 (Electronic)
IS  - 1553-734X (Print)
IS  - 1553-734X (Linking)
VI  - 14
IP  - 11
DP  - 2018 Nov
TI  - Systematically benchmarking peptide-MHC binding predictors: From synthetic to 
      naturally processed epitopes.
PG  - e1006457
LID - 10.1371/journal.pcbi.1006457 [doi]
LID - e1006457
AB  - A number of machine learning-based predictors have been developed for identifying 
      immunogenic T-cell epitopes based on major histocompatibility complex (MHC) class 
      I and II binding affinities. Rationally selecting the most appropriate tool has 
      been complicated by the evolving training data and machine learning methods. 
      Despite the recent advances made in generating high-quality MHC-eluted, naturally 
      processed ligandome, the reliability of new predictors on these epitopes has yet 
      to be evaluated. This study reports the latest benchmarking on an extensive set 
      of MHC-binding predictors by using newly available, untested data of both 
      synthetic and naturally processed epitopes. 32 human leukocyte antigen (HLA) 
      class I and 24 HLA class II alleles are included in the blind test set. 
      Artificial neural network (ANN)-based approaches demonstrated better performance 
      than regression-based machine learning and structural modeling. Among the 18 
      predictors benchmarked, ANN-based mhcflurry and nn_align perform the best for MHC 
      class I 9-mer and class II 15-mer predictions, respectively, on 
      binding/non-binding classification (Area Under Curves = 0.911). NetMHCpan4 also 
      demonstrated comparable predictive power. Our customization of mhcflurry to a 
      pan-HLA predictor has achieved similar accuracy to NetMHCpan. The overall 
      accuracy of these methods are comparable between 9-mer and 10-mer testing data. 
      However, the top methods deliver low correlations between the predicted versus 
      the experimental affinities for strong MHC binders. When used on naturally 
      processed MHC-ligands, tools that have been trained on elution data (NetMHCpan4 
      and MixMHCpred) shows better accuracy than pure binding affinity predictor. The 
      variability of false prediction rate is considerable among HLA types and 
      datasets. Finally, structure-based predictor of Rosetta FlexPepDock is less 
      optimal compared to the machine learning approaches. With our benchmarking of 
      MHC-binding and MHC-elution predictors using a comprehensive metrics, a unbiased 
      view for establishing best practice of T-cell epitope predictions is presented, 
      facilitating future development of methods in immunogenomics.
FAU - Zhao, Weilong
AU  - Zhao W
AUID- ORCID: 0000-0001-8909-5322
AD  - Global Research IT, Merck & Co., Inc., Boston, MA, United States of America.
FAU - Sher, Xinwei
AU  - Sher X
AUID- ORCID: 0000-0003-3977-3695
AD  - Global Research IT, Merck & Co., Inc., Boston, MA, United States of America.
LA  - eng
PT  - Journal Article
PT  - Research Support, Non-U.S. Gov't
DEP - 20181108
PL  - United States
TA  - PLoS Comput Biol
JT  - PLoS computational biology
JID - 101238922
RN  - 0 (Cancer Vaccines)
RN  - 0 (Epitopes, T-Lymphocyte)
RN  - 0 (Histocompatibility Antigens Class I)
RN  - 0 (Histocompatibility Antigens Class II)
RN  - 0 (Ligands)
RN  - 0 (Peptides)
SB  - IM
MH  - Algorithms
MH  - Alleles
MH  - Cancer Vaccines/immunology
MH  - Datasets as Topic
MH  - Epitopes, T-Lymphocyte/chemistry/immunology/*metabolism
MH  - Histocompatibility Antigens Class I/immunology/*metabolism
MH  - Histocompatibility Antigens Class II/immunology/*metabolism
MH  - Humans
MH  - Immunogenicity, Vaccine
MH  - Ligands
MH  - Machine Learning
MH  - Major Histocompatibility Complex/*immunology
MH  - Peptides/chemistry/immunology/*metabolism
MH  - Protein Binding
MH  - Reproducibility of Results
MH  - T-Lymphocytes/immunology
PMC - PMC6224037
COIS- All authors are employed by Merck Co. & Inc.
EDAT- 2018/11/09 06:00
MHDA- 2019/03/15 06:00
PMCR- 2018/11/08
CRDT- 2018/11/09 06:00
PHST- 2018/01/13 00:00 [received]
PHST- 2018/08/22 00:00 [accepted]
PHST- 2018/11/09 06:00 [entrez]
PHST- 2018/11/09 06:00 [pubmed]
PHST- 2019/03/15 06:00 [medline]
PHST- 2018/11/08 00:00 [pmc-release]
AID - PCOMPBIOL-D-18-00068 [pii]
AID - 10.1371/journal.pcbi.1006457 [doi]
PST - epublish
SO  - PLoS Comput Biol. 2018 Nov 8;14(11):e1006457. doi: 10.1371/journal.pcbi.1006457. 
      eCollection 2018 Nov.