Table 1

Problems with genotype patient studies

Area of concernSolution
Is there a clear hypothesis?Use functional data if available
Has population stratification been allowed for?Use individuals drawn from the same population/same racial group
Is the study adequately powered?Calculate in advance the numbers needed in each group depending upon the allelic frequency for the polymorphism(s) under consideration, the number of phenotypes to be studied, the number of polymorphisms to be studied and the likely magnitude of effect of any polymorphism
Is linkage disequilibrium a problem?Study several SNPs within a given gene and consider whether a haplotype analysis may help
Is multiple testing on this population an issue?State the number of studies performed in the population previously and adjust the statistical analysis accordingly