Model for Diversity Analysis of Antigen Receptor Repertoires

TL;DR

This paper introduces a parametric statistical model based on multivariate Poisson-lognormal distribution for analyzing diversity in antigen receptor repertoires, offering insights beyond traditional non-parametric methods.

Contribution

It presents a novel parametric approach that accounts for data under-sampling and enables simultaneous comparison of multiple receptor groups using modern unsupervised learning tools.

Findings

Model fits receptor diversity data well

Outperforms traditional non-parametric methods

Provides insights into clonal size distribution

Abstract

In most of the recent immunological literature the differences across antigen receptor populations are examined via non-parametric statistical measures of species overlap and diversity borrowed from ecological studies. While this approach is robust in a wide range of situations, it seems to provide little insight into the underlying clonal size distribution and the overall mechanism differentiating the receptor populations. As a possible alternative, the current paper presents a parametric method which adjusts for the data under-sampling as well as provides a unifying approach to simultaneous comparison of multiple receptor groups by means of the modern statistical tools of unsupervised learning. The parametric model is based on a flexible multivariate Poisson-lognormal distribution and is seen to be a natural generalization of the univariate Poisson-lognormal models used in ecological studies of biodiversity patterns. The procedure for evaluating model's fit is described along with the public domain software developed to perform the necessary diagnostics. The model-driven analysis is seen to compare favorably vis a vis traditional methods when applied to the data from T-cell receptors in transgenic mice populations.