For a given SNP assumed to be causal under a particular disease model and effect sizes, it is straightforward to calculate the genotype frequencies in cases at that SNP. Case samples are simulated separately by first simulating the genotype of the case at the causative SNP and then working outwards in each direction to construct the haplotypes carrying the alleles simulated at the causative SNP. Loosely, this process will result in oversampling of HapMap chromosomes which carry the risk allele, with the effect dropping off as one moves away (in genetic distance) from the causative locus (see Figures S2, S3, and S4 for examples).
