The sets of results from approaches (1) and (2) were compared to those expected by chance using 100,000 Monte Carlo II simulation trials (below) ([36] and below, and Johnson et al, in preparation). For each of 100,000 simulation trials, a random set of SNPs was chosen by sampling randomly from a list that contained all SNPs studied. The randomly chosen SNPs were considered “pseudopositive” SNPs for that trial. The number of trials for which the results from “pseudopositive” SNPs subjected to our analytical procedure matched or exceeded the results actually observed from the SNPs identified in the current study was tabulated. Empirical p values were calculated by dividing the number of trials for which the observed results were matched or exceeded by the total number of Monte Carlo simulation trials. Similar Monte Carlo III approaches ([36] and below) sampled from a dataset of all gene sequences. These approaches allowed us to generate nominal p values for the observations made for each gene listed in Table 2.
