Dear Reader, Please can you tell me how to use weights with the SNPassoc package function association. association(NEB~recessive(rs178012)*as.factor(GENDER), weights = as.factor(WAGA), data=dat2SNP) SNP: recessive(rs178012 adjusted by: Interaction --------------------- 1 dif lower upper 2 dif lower upper G/G-A/G 1485 1.451 0.03814 0.0000 NA NA 1702 1.854 0.03879 0.4025 0.2946 0.5105 A/A 11 1.091 0.45636 -0.3603 -1.28 0.5599 20 1.900 0.62786 0.4488 -0.2356 1.1333 p interaction: 0.48707 In this example I want to know if there is an interaction between GENDER and the SNP in assocation with NEB, weighted for WAGA. Removal of the weights argument gives exactly the same p.value. association(NEB~recessive(rs178012)*as.factor(GENDER), data=dat2SNP) SNP: recessive(rs178012 adjusted by: Interaction --------------------- 1 dif lower upper 2 dif lower upper G/G-A/G 1485 1.451 0.03814 0.0000 NA NA 1702 1.854 0.03879 0.4025 0.2946 0.5105 A/A 11 1.091 0.45636 -0.3603 -1.28 0.5599 20 1.900 0.62786 0.4488 -0.2356 1.1333 p interaction: 0.48707 Changing the model from recessive into for example dominant gives me other p interactions and p.values, but still the weighting is not giving any differences. Many thanks in advance. Your sincerely -- Thierry [[alternative HTML version deleted]]