The NLR family of receptors plays a key role in the innate immune system of animals, plants and fungi. In the latter two phyla NLRs adapt quickly to ever-changing pathogen-specific invasion markers thanks to their repeat-based architecture, which can pro-duce diversity of recognition epitopes through unequal crossing-over and mutation. Charac-terizing computationally the language of these pathogen recognition receptors can provide insight into the molecular mechanisms of immune response and describe the limits of the pathogen targets that can be recognized. In this work, we model generation and selection of the recognition epitopes as a stochastic string rewriting system with constraints, tuned by analysis of observed evolutionary processes and validated with regard to a large dataset of fungal NLRs. Among others, analyzing the feasible set of solutions revealed that the model explained the i/i + 2 periodicity observed in the repeat number distribution of a family of receptors. In addition, in exploring discrepancies between real and simulated data we discovered an overrepresented pattern which potentially has functional importance. The methodology developed in this work is general and therefore can be applied to any class of amino acid repeats generated by unequal crossing-over for which an equivalent high quality dataset is available.