Detecting and quantifying antibody reactivity in PhIP-Seq data with BEER.

Area of Science:

• Immunology
• Bioinformatics
• Computational Biology

Background:

• Antibodies in peripheral blood are valuable biomarkers for environmental exposures and immune responses.
• Phage Immuno-Precipitation Sequencing (PhIP-Seq) is a high-throughput method for analyzing antibody binding to numerous epitopes.
• Accurate detection of low-affinity antibody responses is critical for understanding immune system dynamics.

Purpose of the Study:

• To introduce BEER (Bayesian Enrichment Estimation in R), a novel software package for quantifying peptide reactivity in PhIP-Seq data.
• To provide a robust tool for analyzing antibody responses at a cohort scale.
• To improve the sensitivity and accuracy of detecting weakly reactive peptides.

Main Methods:

• BEER utilizes a hierarchical Bayesian model to estimate posterior probabilities of peptide reactivity.
• The software calculates fold change estimates to quantify the magnitude of antibody responses.
• BEER also incorporates an alternative peptide reactivity inference method using the edgeR package for comparative analysis.

Main Results:

• BEER accurately quantifies peptide reactivity from PhIP-Seq experiments.
• The Bayesian approach implemented in BEER demonstrates high sensitivity, particularly for weakly reactive peptides.
• While edgeR offers speed improvements, BEER's Bayesian method provides superior sensitivity for subtle antibody responses.

Conclusions:

• BEER is a powerful and accessible R package for analyzing PhIP-Seq data.
• The software enhances the characterization of antibody development and decline by accurately detecting low-affinity responses.
• BEER facilitates large-scale immunological studies by providing precise quantification of antibody reactivity.