Simulating autosomal genotypes with realistic linkage disequilibrium and a spiked-in genetic effect

BMC Bioinformatics

Table 4 Analytic recovery of gene-environment interaction effects for dichotomous and continuous traits in a homogeneous population, based on 1000 simulated studies of 1000 triads (dichotomous) or offspring (continuous)

Phenotype	Allele Frequency	Pathway	Truth or Estimate	Parameter Values^a		Estimated Coverage of Nominal 95% CIs
Phenotype	Allele Frequency	Pathway	Truth or Estimate	Pathway Genetic Effect	GxE Interaction Effect	Pathway Genetic Effect	GxE Interaction Effect
Dichotomous	0.3	1	True	1	1.65
		1	Estimated	1.00 (0.99,1.01)	1.65 (1.61,1.69)	0.953	0.954
		2	True	1	2.71
		2	Estimated	1.00 (0.99,1.02)	2.72 (2.67,2.77)	0.936	0.954
Dichotomous	0.2	1	True	1	1.65
		1	Estimated	1.01 (0.98,1.04)	1.79 (1.67,1.91)	0.959	0.954
		2	True	1	2.71
		2	Estimated	1.01 (0.98,1.04)	3.05 (2.80,3.33)	0.958	0.957
Continuous	0.3	1	True	0.1	0.02
		1	Estimated	0.10 (0.09,0.11)	0.02 (0.00,0.04)	0.944	0.966
		2	True	0.15	0.02
		2	Estimated	0.15 (0.14,0.16)	0.02 (0.00,0.03)	0.948	0.941
Continuous	0.2	1	True	0.1	0.02
		1	Estimated	0.10 (0.08,0.11)	0.03 (−0.01,0.07)	0.945	0.949
		2	True	0.15	0.02
		2	Estimated	0.17 (0.15,0.18)	0.02 (−0.02,0.06)	0.936	0.933

^aParameter values relate to relative risks for dichotomous traits and to mean shifts for quantitative traits. For all models, we assumed that in the absence of either genetic pathway there was no effect of the dichotomous exposure

ISSN: 1471-2105