|
Post by federico on Dec 7, 2021 15:30:42 GMT
I appreciate the question is somewhat redundant, but I have run some analyses with GCTA where the phenotype is 0/1 (basically alive/dead) and I assume GCTA is just fitting a logistic regression with the population structure is corrected with the GRM.
I ask because a colleagues is asking whether GCTA is in fact fitting a linear model with normal errors since the outcome is numeric, but I cannot fathom how that would be. In any case I just need a confirmation on the model used (normal or logistic) when the outcome is 0/1.
|
|