Table 1 Summary of adjusted Embedded Image and slope coefficients from regression of estimated ORs on true or logistic regression ORs from simulated and real-data analysis of type 2 diabetes
NullLogisticEmbedded ImageEmbedded ImageEmbedded ImageEmbedded ImageEmbedded Image rareType 2 diabetes
Embedded Image
Embedded Image0.9870.9990.9930.9960.9860.9830.9970.986
 Slope0.9891.000.9690.9650.9750.9531.020.977
 R2*0.9920.9880.9950.9840.9880.9970.986
 Slope*1.000.9790.9660.9620.9791.010.977
 R2**0.9910.9930.9960.9860.9800.9970.986
 Slope**1.020.9700.9650.9750.9561.020.977
Embedded Image
Embedded Image0.9870.9910.9890.9940.9840.9830.9970.986
 Slope0.9891.050.9980.9841.020.9721.010.977
 R2*0.9910.9890.9940.9820.9890.9970.986
 Slope*1.050.9980.9840.9950.9981.010.977
 R2**0.9850.9890.9940.9840.9790.9970.986
 Slope**1.060.9980.9841.010.9751.020.977
  • Results were generated from regression of transformed ORs from the LMM on estimated ORs from logistic regression for the null and type 2 diabetes results and the true simulated ORs in the logistic and liability threshold simulation scenarios. The * rows correspond to the results from the use of reference allele frequencies for p from the 1000 Genomes Phase 1 Version 3 European sample. The ** rows correspond to the results from the use of the transformations using the SE of the regression coefficient instead of p.