Abstract
We sought a genotype-phenotype association: between single-nucleotide polymorphisms (SNPs) in olfactory receptor (OR) genes from the two largest OR gene clusters and odor-triggered nonallergic vasomotor rhinitis (nVMR). In the initial pedigree screen, using transmission disequilibrium test (TDT) analysis, six SNPs showed "significant" p-values between 0.0449 and 0.0043. In a second case-control population, the previously identified six SNPs did not re-emerge, whereas four new SNPs showed p-values between 0.0490 and 0.0001. Combining both studies, none of the SNPs in the TDT analysis survived the Bonferroni correction. In the population study, one SNP showed an empirical p-value of 0.0066 by shuffling cases and controls with 10(5) replicates; however, the p-value for this SNP was 0.83 in the pedigree study. This study emphasizes that underpowered studies having p-values between < 0.05 and 0.0001 should be regarded as inconclusive and require further replication before concluding the study is "informative." However, we believe that our hypothesis that an association between OR genotypes and the nVMR phenotype remains feasible. Future studies using either a genomewide association study of all OR gene-pseudogene regions throughout the genome--at the current recommended density of 2.5 to 5 kb per tag SNP--or studies incorporating microarray analyses of the entire "OR genome" in well-characterized nVMR patients are required.