Proposed Experiments and Improvements For Maximum Likelihood Models Of Antibody Diversity

Abstract

Previous work has improved the understanding of the statistical features leading to diversity in T-cell receptors [1] [2]. This work has used maximum likelihood models in order to estimate generation probabilities of T-cell receptors from sequence data. It has also employed this type of model to suggest important ideas about the universality of underlying stochastic recombination mechanisms as well as how these mechanisms are selected for during adaption of the immune system. First we overview the core ideas of this previous work and then we look at further work which extend these techniques to investigate how mice immune system diversity differs over age groups [3]. The potential applications of the techniques explored within [3] are then discussed with a new dataset of human subjects at varying states of immune system development.