The rationale that underlies imputation methods is that even though the causal SNP may not have been genotyped in the study at hand, it may have been genotyped in the reference population.