Abstract 2923: Association Mapping Of QT-interval In a 500k Genome-wide Scan: Confirmation Of NOS1AP And Identification Of A Spectrum Of Additional QTLs
Background: The electrocardiographic QT-interval is a normally distributed quantitative trait in the general population, which is associated with sudden cardiac death. Its heritability is between 30% and 50%. In a previous 100k scan we have identified a QTL for QT-interval at the NOS1AP gene which has been replicated in at least five other population samples and explains between 1,5 and 2% of trait variance in both genders. Aim: To map the spectrum of QTLs for QT-interval more comprehensively we undertook a 500k genome-wide association scan in a larger sample of general population individuals from Augsburg, Southern Germany.
Method: From the population-based KORA S4 survey n=1,664 randomly selected individuals were genotyped on Affymetrix 500k arrays. Association was calculated using single and multi-marker tests under additive, dominant and recessive models adjusting for covariates known to explain nongenetic parts of QT variance.
Results: In the 500k dataset the QTL at the NOS1AP gene gave a clear and unmistakably association signal. Overall 60 SNPs throughout a 500kb genomic region were associated with significance levels down to 1e-7.5. In addition we identified 55 additional QTLs between 1 and 5 SNPs in size with significance levels smaller 1e-5. Five of these QTLs successfully replicated in a n=1.200 sample selected for extremely high and low QT intervals (<15% and >85%ile) in order to increase replication power with a minimum in sample size.
Conclusions: The QTL at the NOS1AP gene identified in a 100k scan was now confirmed as the single most significant signal from a 500k-genomewide scan. Its well identifiable signal is due to its high allele frequency (MAF=0.35), 500kb long LD relationship and relatively strong effect size. The newly identified QTLs display a spectrum of different effect sizes, allele frequencies, surrounding LD patterns and coverage by commercial SNP-genotyping arrays, which in combination decreased their detection probability and explains why they went undetected in the previous association scan.