Inferring the immune response from repertoire sequencing.

High-throughput sequencing of B- and T-cell receptors makes it possible to track immune repertoires across time, in different tissues, and in acute and chronic diseases or in healthy individuals. However, quantitative comparison between repertoires is confounded by variability in the read count of each receptor clonotype due to sampling, library preparation, and expression noise. Here, we present a general Bayesian approach to disentangle repertoire variations from these stochastic effects. Using replicate experiments, we first show how to learn the natural variability of read counts by inferring the distributions of clone sizes as well as an explicit noise model relating true frequencies of clones to their read count. We then use that null model as a baseline to infer a model of clonal expansion from two repertoire time points taken before and after an immune challenge. Applying our approach to yellow fever vaccination as a model of acute infection in humans, we identify candidate clones participating in the response. Author summary: High-throughput immune repertoire sequencing (RepSeq) experiments are becoming a common way to study the diversity, structure and composition of lymphocyte repertoires, promising to yield unique insight into individuals' past infection history. However, the analysis of these sequences remains challenging, especially when comparing two different temporal or tissue samples. Here we develop a new theoretical approach and methodology to extract the characteristics of the lymphocyte repertoire response from different samples. The method is specifically tailored to RepSeq experiments and accounts for the multiple sources of noise present in these experiments. Its output provides expansion parameters, as well as a list of potentially responding clonotypes. We apply the method to describe the response to yellow fever vaccine obtained from samples taken at different time points. We also use our results to estimate the diversity and clone size statistics from data. [ABSTRACT FROM AUTHOR]