Benchmarking B cell epitope prediction: underperformance of existing methods

Martin J. Blythe, Darren R. Flower

Research output: Contribution to journalArticle

Abstract

Sequence profiling is used routinely to predict the location of B-cell epitopes. In the postgenomic era, the need for reliable epitope prediction is clear. We assessed 484 amino acid propensity scales in combination with ranges of plotting parameters to examine exhaustively the correlation of peaks and epitope location within 50 proteins mapped for polyclonal responses. After examining more than 106 combinations, we found that even the best set of scales and parameters performed only marginally better than random. Our results confirm the null hypothesis: Single-scale amino acid propensity profiles cannot be used to predict epitope location reliably. The implication for studies using such methods is obvious.
Original languageEnglish
Pages (from-to)246-248
Number of pages3
JournalProtein Science
Volume14
Issue number1
DOIs
Publication statusPublished - Jan 2005

Fingerprint

B-Lymphocyte Epitopes
Benchmarking
Epitopes
Amino Acids
Proteins

Keywords

  • active site/binding site/epitope mapping
  • proteins of the immune system
  • immunological methods
  • epitope prediction

Cite this

Blythe, Martin J. ; Flower, Darren R. / Benchmarking B cell epitope prediction: underperformance of existing methods. In: Protein Science . 2005 ; Vol. 14, No. 1. pp. 246-248.
@article{c079b2dda0ff418dbd35486314a7e1f6,
title = "Benchmarking B cell epitope prediction:: underperformance of existing methods",
abstract = "Sequence profiling is used routinely to predict the location of B-cell epitopes. In the postgenomic era, the need for reliable epitope prediction is clear. We assessed 484 amino acid propensity scales in combination with ranges of plotting parameters to examine exhaustively the correlation of peaks and epitope location within 50 proteins mapped for polyclonal responses. After examining more than 106 combinations, we found that even the best set of scales and parameters performed only marginally better than random. Our results confirm the null hypothesis: Single-scale amino acid propensity profiles cannot be used to predict epitope location reliably. The implication for studies using such methods is obvious.",
keywords = "active site/binding site/epitope mapping , proteins of the immune system , immunological methods , epitope prediction",
author = "Blythe, {Martin J.} and Flower, {Darren R.}",
year = "2005",
month = "1",
doi = "10.1110/ps.041059505",
language = "English",
volume = "14",
pages = "246--248",
journal = "Protein Science",
issn = "0961-8368",
publisher = "Cold Spring Harbor Laboratory Press",
number = "1",

}

Benchmarking B cell epitope prediction: underperformance of existing methods. / Blythe, Martin J.; Flower, Darren R.

In: Protein Science , Vol. 14, No. 1, 01.2005, p. 246-248.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Benchmarking B cell epitope prediction:

T2 - underperformance of existing methods

AU - Blythe, Martin J.

AU - Flower, Darren R.

PY - 2005/1

Y1 - 2005/1

N2 - Sequence profiling is used routinely to predict the location of B-cell epitopes. In the postgenomic era, the need for reliable epitope prediction is clear. We assessed 484 amino acid propensity scales in combination with ranges of plotting parameters to examine exhaustively the correlation of peaks and epitope location within 50 proteins mapped for polyclonal responses. After examining more than 106 combinations, we found that even the best set of scales and parameters performed only marginally better than random. Our results confirm the null hypothesis: Single-scale amino acid propensity profiles cannot be used to predict epitope location reliably. The implication for studies using such methods is obvious.

AB - Sequence profiling is used routinely to predict the location of B-cell epitopes. In the postgenomic era, the need for reliable epitope prediction is clear. We assessed 484 amino acid propensity scales in combination with ranges of plotting parameters to examine exhaustively the correlation of peaks and epitope location within 50 proteins mapped for polyclonal responses. After examining more than 106 combinations, we found that even the best set of scales and parameters performed only marginally better than random. Our results confirm the null hypothesis: Single-scale amino acid propensity profiles cannot be used to predict epitope location reliably. The implication for studies using such methods is obvious.

KW - active site/binding site/epitope mapping

KW - proteins of the immune system

KW - immunological methods

KW - epitope prediction

UR - https://onlinelibrary.wiley.com/doi/full/10.1110/ps.041059505

U2 - 10.1110/ps.041059505

DO - 10.1110/ps.041059505

M3 - Article

VL - 14

SP - 246

EP - 248

JO - Protein Science

JF - Protein Science

SN - 0961-8368

IS - 1

ER -