Control groups like the HapMap use chrom. sample count by which the frequency per genotype is given. Does this imply that you should also use the chrom. sample count of the cases, e.g. 100 patients means 1000 chrom. sample count?
Yes you would need the number of patients because each patient has a genotype. Based on your case-control number of genotypes, then genotype frequencies, you can then compare with that found on the HapMap.