TABLE 4

Powers and type I error rates for simulation 2

Disease-related haplotypeNo. of casesaRelative riskb
Analysis1.01.52.03.0
Nonec200PENHAPLO0.052
500PENHAPLO0.058
1000PENHAPLO0.058
Major200PENHAPLO0.1220.379
Method 40.1130.298
500PENHAPLO0.2920.836
Method 40.2300.734
1000PENHAPLO0.5620.994
Method 40.4670.981
Minor200PENHAPLO0.2460.828
Method 40.2120.712
500PENHAPLO0.7020.999
Method 40.5800.995
1000PENHAPLO0.9761.000
Method 40.9151.000
  • a The number of controls was the same as the number of cases.

  • b The relative risk was defined as the ratio of the penetrance with two disease-related haplotypes to that without any disease-related haplotype. When the simulations of disease-related haplotype were major, the additive risk model was assumed; when they were minor, the dominant model was assumed. The odds ratios for the dominant model corresponded to 2.25 and 3.9.

  • c The null hypothesis.