Inferring the T-cells repertoire dynamics of healthy individuals

Abstract

The adaptive immune system is a diverse ecosystem that responds to pathogens by selecting cells with specific receptors. While clonal expansion in response to particular immune challenges has been extensively studied, we do not know the neutral dynamics that drive the immune system in absence of strong stimuli. Here we learn the parameters that underlie the clonal dynamics of the T-cell repertoire in healthy individuals of different ages, by applying Bayesian inference to longitudinal immune repertoire sequencing (RepSeq) data. Quantifying the experimental noise accurately for a given RepSeq technique allows us to disentangle real changes in clonal frequencies from noise. We find that the data are consistent with clone sizes following a geometric Brownian motion, and show that its predicted steady state is in quantitative agreement with the observed power-law behaviour of the clone-size distribution. The inferred turnover time scale of the repertoire increases substantially with patient age, and depends on the clone size in some individuals.

NoisET: Noise Learning and Expansion Detection of T-Cell Receptors

“…Using NoisET, one can also generate in-silico RepSeq samples, and their neutral dynamics following the stochastic population dynamics developed in ref and in ref . The function takes as input the noise model method (negative binomial or Poisson), the noise model parameters at both time points, the number of reads at both time points, the duration of the simulations, and the values of τ and θ describing the global stochastic population dynamics.…”

“…More broadly, the noise inference using the first module of NoisET has also been used to learn the natural dynamics of T cell repertoires in the absence of specific antigenic stimulation . For all individuals studied we found a universal constraint on the dynamics, which translated into a susceptible dynamical system that can easily maintain a large number of diverse cells.…”

“…So far, NoisET has been used to study the short time scale dynamics for acute infections, but could also be used to compare bulk repertoires with selected repertoires derived from functional or cultured assays . For longer time scales, the dynamics of lymphocyte populations should be modeled to best describe slow global repertoire changes that cannot be attributed to a single stimulus. , …”

“…21 For longer time scales, the dynamics of lymphocyte populations should be modeled to best describe slow global repertoire changes that cannot be attributed to a single stimulus. 27,28 The Bayesian approach encoded in NoisET results in a more reliable way to account for uncertainty than statistical estimates that are also less interpretable. The detection of responding clones based on the fold change of empirical abundances was not optimal without a robust interpretation of the details of the noise model.…”

“…As mentioned in section 4 the noise learning feature of NoisET has also been used to learn the natural dynamics of TCR repertoires based on gDNA and cDNA data in the absence of direct antigenic stimulation. 28 This study considered the TCRβ repertoires of 9 people and showed that the dynamics of all people, regardless of age is constrained by the power law exponent of the frequency distribution. The exponent itself is given by the ratio of the deterministic turnover time scale and the stochastic noise time scale.…”