chips, yield for the two markers used for the surrogate (rs12477914 and rs1366840) was at least 98% and, in the controls of each group, neither marker deviated significantly from Hardy–Weinberg equilibrium (P > 0.05). For both the markers used, when the same study group was typed on more than one chip, there was no significant difference in allele frequency between the chips in either cases or controls (P > 0.05).
