Exact simulation for disease status
Samples will be simulated with exactly the same disease status as specified in the six column of the proband file. A retrospective likelihood P(G|X), where X is a vector of disease status and G is a vector of disease haplotypes for all family members, will be calculated. The likelihood is efficiently approximated when the probability of affection status for family members that have not been simulated given the haplotypes and affection statuses for members that have been simulated is calculated within nuclear families. Our simulation results suggested that this option should be used when the disease prevalence is < 0.05.
For more details of the algorithm, please refer to the paper:
Yao PJ and Chung RH. SeqSIMLA2_exact: Simulate multiple disease sites in large pedigrees with given disease status for disease with low prevalence. Submitted.