Table 5.
Summary statistics for parameter estimates derived from case-parent triad data with genotypes missing for one parent
| β1 = 0.405 | β2 = 0.405 | β3 = 0.405 | β5 = 1.100 | |
|---|---|---|---|---|
| Gene-environment interaction model | ||||
| Full | 0.391 (0.281, 0.280) | 0.355 (0.213, 0.218) | 0.350 (0.216, 0.218) | 0.828 (0.469, 0.464) |
| Complete | 0.404 (0.425, 0.455) | 0.350 (0.346, 0.349) | 0.350 (0.351, 0.350) | 1.060 (0.751, 0.767) |
| Imputation | 0.398 (0.303, 0.309) | 0.361 (0.244, 0.240) | 0.357 (0.242, 0.240) | 0.838 (0.442, 0.524) |
| Gene-gene interaction model | ||||
| Full | 0.378 (0.255, 0.251) | 0.379 (0.245, 0.251) | 0.354 (0.225, 0.218) | 0.852 (0.383, 0.369) |
| Complete | 0.405 (0.409, 0.404) | 0.401 (0.407, 0.405) | 0.355 (0.321, 0.349) | 0.852 (0.383, 0.369) |
| Imputation | 0.384 (0.276, 0.274) | 0.386 (0.275, 0.276) | 0.356 (0.255, 0.240) | 0.850 (0.346, 0.399) |
The results are based on 1000 simulated datasets, each consisting of 100 triads and 150 dyads. ‘Full’ data analyses are based on all 250 triads with no missing genotypes; ‘Complete’ data analyses use only the 100 triads with genotypes available for both parents; ‘Imputation’ data analyses are based on 100 triads and 150 dyads by using a multiple imputation method. Each entry lists the mean estimate (standard deviation of the estimates, mean of the estimated standard errors) over the 1000 simulated datasets.
The main effect of the environmental covariate β4 is unestimable from triads-only and thus omitted.