Fig 1.
Patterns of frequency dependence and heterozygote advantage that emerged from the model.
(A) Comparison of pathogen recognition abilities between heterozygotes and homozygotes, depicted as averages across the last 6000 generations from 10 independent runs. Mutation rates for (A-D) are given in the legends; host population size N = 5000 individuals. (B-D) Relationship between the expected proportion of pathogens recognized by MHC molecules (taken at generation 2000, i.e. at the time when allele numbers stabilized) and allele frequency is shown for three scenarios: Red Queen process (RQ), heterozygote advantage (HA), and both (HA+RQ). Points represent MHC alleles; colors denote pathogen mutation rates (see the legend). Lines represent the statistically significant trends in frequency dependence (see Table A in S2 Appendix).
Fig 2.
Population size and MHC polymorphism.
Impact of population size on the maintenance of MHC polymorphism (upper panels) and heterozygosity (lower panels) under three simulated scenarios–heterozygote advantage (HA), Red Queen (RQ) dynamics, and both (HA+RQ) (see legend)–with three pathogen mutation rates (given above the panels). Data points indicate the mean number of MHC alleles (A-C) and mean heterozygosity (D-F) and are calculated across the last 6000 generations of 10 replicate simulations. Error bars indicate bootstrapped 95% confidence intervals.
Fig 3.
Presentation ability and probability of recruitment of a new mutant MHC allele.
(A-C) Data points indicate the probability that a mutant allele will stay in the population for at least 10 host generations (cut-off points longer or shorter than 10 generations (e.g., 2 or 20) showed similar patterns). (D-F) Data points indicate the relative pathogen-recognition ability of a mutant MHC allele relative to the immunocompetence of resident alleles. The average immunocompetence of resident alleles was measured as the number of pathogens recognized in a given generation, weighted by the frequency of each resident allele. Note that a presentation spectrum equal to 1 indicates the threshold at which a mutant allele is able to present, on average, the same proportion of pathogens as resident alleles are. (A-F) Characteristics were calculated across the last 6000 generations of 10 replicates. Scenario labels: HA–heterozygote advantage, RQ–Red Queen, HA+RQ–both. Pathogen mutation rates are given above the panels.
Fig 4.
Frequency dynamics and presentation abilities of mutant alleles in the initial period following a mutation’s appearance.
Panels in the upper row show averaged allele frequency dynamics over 25 generations after an allele mutation occurred under three mutation rates (see legend). Panels in the bottom row show the dynamics of median mutant allele immunocompetence in relation to the immunocompetence of resident alleles, measured as the number of pathogens presented in a given generation (weighted by the frequency of resident alleles). The measure was calculated across the last 6000 generations of 10 replicate simulations with the host population size N = 5000. Note that a presentation spectrum equal to 1 in the bottom row panels indicate the threshold at which a mutant allele is able to recognize, on average, the same proportion of pathogens as the resident alleles. For presentation purposes the scale on the Y-axis in the bottom row panels was log-transformed. Scenario labels: HA–heterozygote advantage, RQ–Red Queen, HA+RQ–both.
Fig 5.
Persistence of MHC allelic lineages.
(A) Fraction of allele pairs with no common ancestor within 40 000 generations (red, green, and blue symbols) and 250 000 generations (gray symbols, run only for the highest mutation rate of 5e-3) and (B) coalescence times for studied scenarios given in the legend (HA–heterozygote advantage, RQ–Red Queen, HA+RQ–both).