First, you need to build your genotype database in a numeric format (i.e. TT=0, TC=1, CC=2, control=0, patient=1, etc) and then use a logistic regression analysis with odd ratios. Just remember that response variable has to be dichotomous (i.e. control=0, patient=1).