We applied our two approaches to the first 50 replicates for which SOLAR obtained a convergent solution. The significant results (defined as p < 10−4 from either method) are reported in Table 1. After receiving the simulation parameters, we divided these into true positives and false positives. For the traditional linkage analysis approach, we list the median p-value from the individual replicates, the range of p-values, and the p-value obtained by combining the data from the families across the 50 replicates (i.e., an analysis based on a sample 50 times the size of a single replicate, N = 20,000). The novel repeated measures analysis (performed using the GLMM) is an analysis based on only the individuals from a single replicate (N = 400), but it uses the information from 50 repeated measures of each trait to provide a more accurate estimate of the total genetic liability for each individual.
