n | H0 : β1 = 0 | H0 : β5 = 0 | ||||||||
---|---|---|---|---|---|---|---|---|---|---|
OLR (16) † | AQTDTM (17) | FBAT (18) | OLMM (19) | ALMM (20) | ||||||
HP | SP | HP | SP | HP | SP | HP | SP | HP | SP | |
100 | 0.108 | - | 0.074 | 0.074 | 0.075 | 0.072 | 0.146 | - | 0.116 | 0.09 |
200 | 0.155 | - | 0.104 | 0.098 | 0.103 | 0.101 | 0.259 | - | 0.166 | 0.118 |
300 | 0.237 | - | 0.161 | 0.128 | 0.158 | 0.128 | 0.359 | - | 0.236 | 0.164 |
400 | 0.255 | - | 0.182 | 0.143 | 0.178 | 0.144 | 0.469 | - | 0.295 | 0.18 |
500 | 0.357 | - | 0.226 | 0.176 | 0.227 | 0.175 | 0.552 | - | 0.342 | 0.221 |
600 | 0.411 | - | 0.275 | 0.206 | 0.278 | 0.275 | 0.605 | - | 0.417 | 0.255 |
Table S4b.Empirical power results for gene-environment effect assessment comparing ordinary statistical methods (OLR and OLMM) with family-based methods (AQTDT M , QBAT-I and ALMM) under homogeneous (HP) and stratified (SP) populations. Each time, n cases were simulated with parameters β 3 = α 7 = 1. Simulations are based on the recessive genetic model. † = number that identifies each model in the paper. | ||||||||||
n | H 0 : β 3 = 0 | H 0 : β 7 = 0 | ||||||||
OLR (21) † | AQTDT M (22) | QBAT-I | OLMM (23) | ALMM (24) | ||||||
HP | AP | HP | AP | HP | AP | HP | AP | HP | AP | |
100 | 0.099 | - | 0.067 | 0.078 | 0.070 | 0.065 | 0.151 | - | 0.118 | 0.098 |
200 | 0.160 | - | 0.127 | 0.111 | 0.119 | 0.109 | 0.252 | - | 0.180 | 0.144 |
300 | 0.220 | - | 0.150 | 0.124 | 0.130 | 0.120 | 0.374 | - | 0.219 | 0.155 |
400 | 0.296 | - | 0.215 | 0.132 | 0.158 | 0.129 | 0.466 | 0.302 | 0.172 | |
500 | 0.345 | - | 0.241 | 0.160 | 0.182 | 0.161 | 0.535 | - | 0.342 | 0.216 |
600 | 0.422 | - | 0.252 | 0.192 | 0.201 | 0.190 | 0.633 | - | 0.402 | 0.260 |