I have read in the issues threads that the BETAs might be biased in some circumstances, especially for low frequency variants.
I wonder if the direction (positive or negative) of the BETA can be trusted? For a binary variable, would it be correct to say that if the BETA value is positive for an association between a SNP and a disease, the SNP increase the risk of disease?
Hi,
I have read in the issues threads that the BETAs might be biased in some circumstances, especially for low frequency variants. I wonder if the direction (positive or negative) of the BETA can be trusted? For a binary variable, would it be correct to say that if the BETA value is positive for an association between a SNP and a disease, the SNP increase the risk of disease?