Automated High-Throughput Mapping of Linear B-Cell Epitopes Using a Statistical Analysis of High-Density Peptide Microarray Data

Abstract

Detailed information of antibodies’ specificity is often missing or inadequate even for continuous (i.e., linear) epitopes. Recent developments in peptide microarray technology has enabled the synthesis of up to two million peptides per array thereby allowing linear peptide epitopes to be examined by a systematic amino acid substitution and positional scanning approach. This kind of analysis generates a very large body of data, which needs to be analyzed and interpreted in a robust and automated manner. Here, we describe a rational systematic approach to define linear antibody epitopes using ANOVA statistics to identify not only significant but also important residues involved in antibody recognition. This statistical approach can be used to perform a comprehensive linear epitope discovery. For polyclonal antibodies, this could be extended to entire proteins pinpointing critical residues for each epitope. We argue that the ANOVA analysis levels out issues of unknown peptide concentration/quality and unknown antibody titers leading to identification of epitopes that otherwise would be neglected if the evaluation was based merely on signal strength.