Abstract 403: A Genome-wide Association Study to Identify Common Sequence Variants Related to Five Circulating Hemostatic Factor Proteins: The Framingham Heart Study
Background: The circulating level of a hemostatic factor protein may be influenced by single nucleotide polymorphisms (SNPs) proximate to the gene encoding the protein (a cis-acting SNP) or by SNPs at loci that are physically distant (a trans-acting SNP). The proportion of circulating biomarker variation explained by trans-acting SNPs is unknown. The identification of such SNPs may uncover novel pathways involved in the regulation of plasma biomarkers.
Methods: In 1071 related Framingham Heart Study participants (mean age 52 years, 51% women), we analyzed the association of 110,919 SNPs (Affymetrix 100K GeneChip) to five plasma hemostatic factor traits: factor VII antigen (FVII), von Willebrand factor (vWF), plasminogen activator inhibitor-1 antigen (PAI-1), fibrinogen (FG), and tissue plasminogen activator antigen (tPA). We used generalized estimating equations to test the association between SNP genotype and multivariable-adjusted residual trait values in an additive genetic model. SNPs with a minor allele frequency<5%, genotype call rate<80%, or Hardy-Weinberg equilibrium P<0.001 were excluded.
Results: SNPs with the lowest P values for each of the five traits are shown in the Table⇓. Of 84,633 tests for the multivariable-adjusted FVII trait, the strongest association was for SNP1 (P=5*10−16). This SNP (rs561421) is highly correlated (r2 = 1) to the Arg/Gln FVII SNP previously shown to explain 9% of the variance in circulating FVII level. Except for FVII, the most strongly associated SNPs were all trans-acting. For each multivariable-adjusted trait, the number of SNPs with association P<10−5 ranged from one to five and with P<10−4, from 14 to 35.
Conclusions: Using a 100K genome-wide association scan, we verified a previously reported cis-acting genetic association and we identified several novel trans-acting gene loci that are associated with hemostatic factor level. Efforts to replicate the strongest associations in a separate sample are in progress