Table 2.

Sensitivity, specificity, and positive and negative predictive values of each tested approach

	A	B	C	D	E	F	G	H
Sensitivity	0.066	0.066	0.133	0.600	0.066	0.133	0.133	0.133
Specificity	0.987	0.984	0.978	0.761	0.994	0.993	0.979	0.978
Positive predictive value	0.020	0.025	0.029	0.001	0.052	0.095	0.029	0.028
Negative predictive value	0.995	0.995	0.995	0.997	0.995	0.995	0.995	0.995

Each column represents a tested approach: (A) Using the absolute sum of rare variants as a covariate in the polygenic model, using the entire cohort. Selected genes were those that significantly reduced the trait heritability. (B) Using the absolute sum of nonsynonymous rare variants as a covariate in the polygenic model, using the entire cohort. Selected genes were those that significantly reduced the trait heritability. (C) Counting the absolute number of minor alleles in each common variant from a gene and using the gene as a covariate in the polygenic model. (D) Using the number of minor alleles present in common variants as a covariate in the polygenic model, using each family separately. Selected variants were those that significantly reduced the trait heritability in at least one family. (E) QTDT results using the common variants and a linear model of association. (F) QTDT results using common variants and the variance components model. (G) QTDT results using rare variants and a linear model of association. (H) QTDT results using rare variants and the variance components model.