Rare event simulation in immune biology: Models of negative selection in T-cell maturation

We present a probabilistic T-cell model that includes negative selection and takes contrasting models of tissue-restricted antigen (TRA) expression in the thymus into account. We start from the basic model of van den Berg, Rand, and Burroughs (2001) and include negative selection via individual-based T-cell modelling, in which each T-cell is defined by its stimulation rates to all relevant self antigens. We present a simulation approach based on partial tilting of the stimulation rates recognized by a single T-cell. We investigate the effects of negative selection for diverging modes of thymic antigen presentation, namely arbitrary TRA presentation, and more or less strict emulation of tissue-specific cell lines. We observe that negative selection leads to truncation of the tail of the distribution of the stimulation rates mature T-cells receive from self antigens, i.e., the self background is reduced. This increases the activation probabilities of single T-cells in the presence of non-self antigens.