We estimated the genetic relationships among all of the 4,259 individuals in the combined dataset by equation [6]. The estimated relationships (off-diagonal elements of the relationship matrix) ranged from −0.024 to 0.585, suggesting that some close relatives still remained. The mean of genetic relationships of “unrelated” individuals should be close to zero, so the lower-bound of the range can be roughly regarded as the maximum deviation of an estimate from the mean. We estimated the two-tailed 95% confidence interval of relationships (adjusted for multiple tests by Bonferroni correction) to be from about −0.027 to 0.027. Therefore, to avoid having any close relatives in the data, we chose a cut-off value of 0.025 and selectively excluded one of any pair of individuals with an estimated relationship > 0.025 to maximize the remaining sample size. We excluded 287 individuals from the adult dataset and 47 individuals from the adolescent dataset. A total of 3,248 “unrelated” adults and 677 “unrelated” adolescents, with a combined dataset of 3,925 “unrelated” individuals, was retained for analysis.
