2009 Abstract of Approved DNA Project

2009 Abstract of Approved DNA Project

Impact of Common Variation in Candidate Genes on Metabolomic Predictors of Diabetes

PI: Robert Gerszten, MD

Description
This application seeks to supplement ongoing metabolomics research in the Framingham Offspring Study (FOS) with targeted genotyping of candidate genes to identify common genetic variants associated with metabolite levels.

Background
Through metabolite profiling of plasma from subjects in the FOS, we have identified several neutral amino acids that predict future diabetes. These results raise the possibility that inherited differences, as reflected by common variants in candidate genes, might influence the distribution of these metabolites across subjects. Indeed, recent work by others has demonstrated that common variants in key metabolic genes are associated with differential metabolite profiles in human plasma [1, 2]. We hypothesize that common variants in selected genes involved in amino acid metabolism, or known to be associated with diabetes or hyperlipidemia, will be associated with plasma levels of our diabetes-predicting amino acids and other metabolites.

Aim
We propose to evaluate common variants at the candidate loci detailed herein (see Table at end of application) for association with plasma levels of the diabetes-predicting amino acids and other small molecules identified by our group.

Methods
We will perform genotyping on all subjects who attended exam 5 in the Framingham Offspring Study, for whom DNA specimens are available with appropriate participant consent for DNA use.

The NHLBI Human Exome Project

PI: Larry Atwood, PhD

Description
We propose to sequence the exome in every eligible Framingham participant.

If the consortium is unable to sequence all Framingham participants then we propose to genotype those not sequenced for variants identified by the sequencing.

Background
In May, the Framingham investigators joined a Consortium that responded to the NIH Grand Opportunity (GO) that was part of the federal stimulus package. The GO initiative announcement called for “in-depth identification of disease-causing genetic variants using large-scale sequencing of human subjects from longitudinal population cohorts.” The announcement required that the cohort be well-phenotyped for heart, lung and blood traits and comprise multiple ethnic groups.

Aim
The aim is to sequence the exome in all six NHLBI cohorts, conduct follow-up genotyping of variants discovered by sequencing, and to analyze the variants identified for association with a wide range of heart, lung, and blood phenotypes.

Methods
Sequencing will performed at sequencing centers selected by NHLBI in an application process separate from this one. Framingham will provide 5ug of DNA per participant to a sequencing center designated by NHLBI. For genotyping, NHLBI will designate a genotyping center to which Framingham will ship DNA.

The Role of KCNH2/hERG1 K897T SNP in the Susceptibility to Diabetes Mellitus Traits

PI: Stavros Garantziotis, MD

Description
The human ether-a-go-go-related gene, hERG1 (KCNH2), encodes Kv11.1 potassium channels that are essential for rhythmic excitability of cardiac muscle. SNPs in the hERG1 gene are responsible for LQT2 in humans and increase the risk of fatal cardiac arrhythmias. K897T/rs1805123 is one of the most common polymorphisms in the human Kv11.1 channel protein. We discovered recently that K897T creates a new phosphorylation site for the Akt protein kinase, which inhibits channel activity. Consequently, hormonal signaling through the phosphatidy–inositol 3 kinase (PI3K), which increases Akt activity, is predicted to prolong ventricular action potentials in subjects with T897.Kv11.1 potassium channels are also present in pancreatic beta cells, where they regulate insulin secretion. Insulin release inhibits further secretion by signaling through PI3K. Therefore, we hypothesize that individuals with the T897 polymorphism in hERG1 will react to an acute glucose load with an overshooting insulin secretory response. Over time, such individuals could develop a hyperinsulinemic state, and, eventually, insulin resistance. In that case they might exhibit a higher lifetime risk for type II diabetes mellitus. Our hypothesis is supported by a recent publication suggesting that SNPs in another potassium channel, KCNQ1, is associated with susceptibility to DM II, as predicted by our work.

Background
The main objective of this study is to characterize the effect of a common SNP in the human potassium channel gene hERG1 on diabetic traits. Type 2 diabetes has a strong genetic component, with approx 70% concordance in monozygotic twins compared to 20-30% in dizygotic twins. The human ether-a-go-go-related gene, hERG1 (KCNH2), encodes Kv11.1 potassium channels that are essential for rhythmic excitability of cardiac muscle (Sanguinetti et al., 2006). Single nucleotide polymorphisms (SNPs) in the hERG1 gene that reduce Kv11.1 activity are responsible for LQT2 in humans (Curran et al., 2005) and increase the risk of fatal cardiac arrhythmias (Sauer et al., 2007). K897T/rs1805123 is one of the most common polymorphisms in the human Kv11.1 channel protein (Newton-Cheh et al., 2007). We discovered recently that the hERG1 K897T polymorphism creates a new phosphorylation site for the Akt protein kinase on the hERG1 channel proteins, which inhibits their activity (Gentile et al., 2008). Consequently, hormonal signaling through the phosphatidy–inositol 3 kinase (PI3K), which increases Akt activity, is predicted to prolong ventricular action potentials in subjects with the T897 polymorphism. hERG1 channels are also present in pancreatic beta cells, where they regulate insulin secretion (Rosati et al., 2000). Insulin release inhibits further secretion by signaling through PI3K (Manning et al., 2007). Therefore, we hypothesize that individuals with the T897 polymorphism in hERG1 will react to an acute glucose load with an overshooting insulin secretory response. Over time, such individuals could develop a hyperinsulinemic state, and, eventually, insulin resistance. In that case they might exhibit a higher lifetime risk for type II diabetes mellitus. Our hypothesis is supported by a recent publication suggesting that SNPs in another potassium channel, KCNQ1, is associated with susceptibility to DM II (Yasuda et al., 2008), as predicted by our work.

Aim
The Framingham consortium has genotyped approximately 2,500 individuals for the rs1805123/K897T SNP. The results of this genotyping effort are available for our analysis by our collaborator Dr Newton-Cheh, and will be supplied to Dr Dupuis for her analysis. Virtually all of these individuals have also been phenotyped with regard to diabetes-related traits (Meigs et al., 2007). We would like to take advantage of this pre-existing database in support of our basic research data. The phenotypic endpoints we propose would be: 1) Incident type II diabetes; 2) Fasting plasma glucose; 3) Hemoglobin A1c; 4) Fasting insulin; 5) Homeostasis model insulin resistance (HOMA-IR); and 6) 0–120 min insulin sensitivity. At this point, the plan is that genotype – phenotype analysis will be conducted at BUMC by FHS investigators (Dr Josee Dupuis and Dr James Meigs) and aggregate results will be shared with Dr Stavros Garantziotis. However, we may request access to raw data in the future. An IRB proposal in currently being submitted at the NIEHS. If positive results are obtained from this study, we will expand it by investigating additional hERG1 SNPs, and possibly SNPs in other potassium channels. This analysis will therefore prove informative in designing future research. Meigs, J.B., et al., Genome-wide association with diabetes-related traits in the Framingham Heart Study. BMC Med Genet, 2007. 8 Suppl 1: p. S16.

Methods
This will be a cross-sectional gene association study with a specific candidate SNP and multiple quantitative phenotypes. We will also utilize the prospective nature of the FHS to analyze time-to-event data for a qualitative trait.

Genome Variation of 10q11 near CXCL12 and 1q41 in MIA3, in addition to Chromosome 6p24 in PHACTR1, in Framingham Heart Study Participants with Extreme Coronary Artery Disease Phenotypes

PI: Christopher O’Donnell, MD, MPH

Description
We have recently completed a genomewide association study for coronary artery calcium (CAC) and a separate genomewide association study for early onset myocardial infarction in which we discovered a strong association between SNPs on chromosomes 9p21 and 6p24 and both CAC and myocardial infarction. We have completed or are undertaking sequencing of rare variants in 9p21 and we plan to conduct sequencing of 6p24 in Framingham Heart Study individuals with CHD and/or high levels of coronary artery calcium (CAC). DNA from 188 Framingham participants with early onset CHD and/or high levels of age-adjusted CAC and from 94 participants with no CHD or CAC will be extensively resequenced for exons and introns in the PHACTR1 gene. In this addendum, we propose to sequence two other novel loci that were recently reported to be strongly associated with MI and/or CAC in the MIGen Consortium and the Framingham Heart Study: 10q11 (CXCL12) and 1q41 (MIA3). We will conduct statistical tests including Chi square tests to compare the prevalence of sequence variants between the two extreme phenotype groups. We will further genotype top SNPs identified from resequencing in 7,357 FHS Offspring and Gen3 subjects, to examine for associations between SNPs and CAC and CHD/CVD, as well as, secondarily, with traditional risk factors and selected biomarkers, other subclinical vascular and ventricular disease measures. We aim to identify putativ functional variants that alter subclinical and clinical CHD susceptibility in humans in order to identify new therapeutic targets for prevention and treatment of CHD.

Genome Variation in Chromosome 6 PHACTR1 in Framingham Heart Study participants with extreme coronary artery disease phenotypes

PI: Christopher O’Donnell, MD, MPH

We have recently completed a genomewide association study for coronary artery calcium (CAC) and a separate genomewide association study for early onset myocardial infarction in which we discovered a strong association between SNPs on chromosome 6p24 and both CAC and myocardial infarction. This project aims to examine the prevalence of rare variants in 9p21 in Framingham Heart Study individuals with CHD and/or high levels of coronary artery calcium (CAC). DNA from 188 Framingham participants with early onset CHD and/or high levels of age-adjusted CAC and from 94 participants with no CHD or CAC will be extensively resequenced for exons and introns in the PHACTR1 gene. Chi square tests will be used to compare the prevalence of sequence variants between the two extreme phenotype groups. We further propose to genotype top SNPs identified from resequencing in 7,357 FHS Offspring and Gen3 subjects, and to examine for associations between SNPs and CAC and CHD/CVD, as well as, secondarily, with traditional risk factors and selected biomarkers, other subclinical vascular and ventricular disease measures. We aim to identify putative functional variants that alter subclinical and clinical CHD susceptibility in humans in order to identify new therapeutic targets for prevention and treatment of CHD.

Genome Variation in Chromosome 9 in Framingham Heart Study Participants with Extreme Coronary Artery Disease Phenotypes

PI: Christopher O’Donnell, MD, MPH

Recent genome-wide association studies have discovered a strong association between SNPs on chromosome 9p21 and coronary heart disease (CHD). In Framingham Heart Study subjects, we observed associations of SNPs in 9p21 with phenotypes including CHD and high levels of coronary artery calcium, as well as insulin resistance and lipid metabolism traits. This project aims to extend the findings from a sequencing project that identified new variants in 9p21 in Framingham Heart Study individuals with CHD and/or high levels of coronary artery calcium (CAC). DNA from 188 Framingham participants with early onset CHD and/or high levels of age-adjusted CAC and from 94 participants with no CHD or CAC was extensively re-sequenced for missense and nonsense variants in CDKN2A and CDKN2B and in particular for variation in a large gene-free sequence in 9p21. Re-sequencing will be performed by the NHLBI Re-sequencing and Genotyping Service. In this project, we will genotype variants identified from the re-sequencing project in 7,353 men and women from the Gen3 and Offspring cohorts. We will examine associations with multiple phenotypes, including traditional risk factors and selected biomarkers, subclinical vascular and ventricular disease measures, and clinically apparent CVD outcomes, we will compare results with those for the 9p21 region using the 2.5 million SNPs imputed from HapMap and now being analyzed in conjunction with the Framingham Heart Study SNP Health Association Resource (SHARe). We will submit the completed genotypes into dbGaP according to existing policies, to make this resource available to the approved users of the SHARe database.

Genetics of Adult Height in the Framingham Heart Study

PI: Joel Hirschhorn, MD, PhD

Goals
To sequence genes near variants that have been associated with height, using patients at the extremes of the height distribution in the Framingham Heart Study, to determine whether rarer more penetrant variants may contribute to variation in adult height.

Phenotypes
Height

Laboratory Methods
We will carry out next generation sequencing using Illumina Genome Analyzer machines, and using protocols developed at the Broad Institute.

Analytical Approach
We will screen for genes with an excess of missense mutations (particularly those predicted to be “damaging”) in individuals at one end of the height distribution, using multiple samples and thresholds that provide reasonable power while keeping the false discovery rate below 5%.

Sequencing of Genes near Loci Associated with Fatty Liver Disease

PI: Joel Hirschhorn, MD, PhD

Goals
To sequence genes near loci that have been associated with fatty liver in the extremes of fatty liver in the Framingham Heart Study to determine whether new rarer variants exist that may contribute to variation in this trait.

Phenotypes
Liver fat

Laboratory Methods
We will carry out next generation sequencing using Illumina Genome Analyzer machines using the protocols already worked out at the Broad Institute.

Analytical Approach
We will screen for genes with more missense mutations in individuals with fatty liver than in controls in multiple cohorts, using thresholds that provide reasonable power while keeping the false discovery rate below 5%.

Sequencing of Genes near Loci Associated with Obesity

PI: Joel Hirschhorn, MD, PhD

Goals
To sequence genes near loci that have been associated with obesity in obese and lean people in the Framingham Heart Study to determine whether new rarer variants exist that may contribute to variation in this trait.

Phenotypes
Body mass index.

Laboratory Methods
We will carry out next generation sequencing using Illumina Genome Analyzer machines using the protocols already worked out at the Broad Institute.

Analytical Approach
We will screen for genes with more missense mutations in individuals with obesity than in lean individuals, using thresholds that provide reasonable power while keeping the false discovery rate below 5%.

Request for Additional DNA and Genes for Sequence Analysis

PI: Christine Seidman, MD

Description
The focus of this project is to define the spectrum of allelic variation in cardiovascular disease genes in unrelated subjects enrolled in the FHS (n= 1800) and to determine their impact singly and in combination on cardiovascular risk factors. We will sequence genes previously demonstrated to cause monogenic traits and related genes that participate in these pathways. Through this amendment we request additional DNA so that new sequencing platforms can be used to interrogate a large number of candidate genes for single nucleotide polymorphisms and for copy number variations.

Aim
To sequence cardiovascular disease genes in unrelated subjects enrolled in the FHS (n= 1800) to define the spectrum of allelic variation (including SNPs and CNVs) and to determine their impact singly and in combination on cardiovascular risk factors.

Methods
We will construct a genomic library from each subject, capture the cardiovascular subgenome (e.g., cardiovascular genes proposed for analyses), and to sequence the cardiovascular subgenome using Illumina platform. Analyses of exons and flanking intron sequences and copy numbers in the cardiovascular subgenome will be studied to determine:

  1. The frequency and distribution of rare and common sequence variants in candidate genes
  2. The association of rare and common sequence variants, individually and in aggregate, with cardiovascular disease risk factors. We will consider both dichotomous traits (e.g., hypertension) and quantitative traits (e.g., systolic blood pressure).
  3. The relationships of rare and common sequence variants, individually and in aggregate, to subclinical phenotypes including left ventricular hypertrophy and dilation, carotid intimal wall thickness (a surrogate marker for atherosclerosis) and clinical CVD (including coronary heart disease, myocardial infarction, heart failure, ischemic stroke).
  4. Potentially important gene by gene and gene by environment interactions in the associations of sequence variants with risk factor phenotypes, subclinical phenotypes and clinical CVD.

Atrial Fibrillation Risk Prediction, Clinical, Biological and Genetic Markers

PI: Emelia Benjamin, MD, ScM

Description
Goal is to markedly improve ability to predict the onset of atrial fibrillation to provide opportunities for preventing atrial fibrillation in the community, and to identify high risk individuals to target for prevention trials.

  1. To validate, recalibrate and potentially modify the FRS AF a clinical risk prediction model for atrial fibrillation in diverse communities using widely available clinical factors.
  2. To test whether biomarkers, specifically BNP or CRP enhance clinical risk prediction (discrimination, calibration, reclassification) of AF when added to standard clinical factors.
  3. To examine whether a genotype score of top SNPs from CHARGE AF improves risk prediction (discrimination, calibration, reclassification) of AF over and above standard clinical factors.

Background
Atrial fibrillation (AF) is an important public health problem. The prevalence of AF doubles for each advancing decade of life affecting more than 10% of individuals over the age of 80. 1 The lifetime risk of AF is about 25%. 2 In addition, with the aging of the population, and increased survival with cardiovascular disease (CVD) the prevalence of AF is increasing over time. 3 It is estimated that the prevalence of AF in the U.S. alone will rise from 2.3 million in 2001 to between 6 and 15 million in 2050. 4,5 Furthermore, AF is a major source of cardiovascular disease (CVD) morbidity and mortality. It is associated with a five-fold increased stroke risk,6 a doubling in dementia risk,7 a tripling in congestive heart failure (CHF) risk,8 and 1-1/2 to 2-fold adjusted increased odds of death.9 The risk factors for AF are multi-factorial and include CVD and its risk factors.1,10,11

Aim

  1. To validate, recalibrate and potentially modify the FHS AF clinical risk prediction model in diverse communities using widely available clinical factors.
  2. To test whether biomarkers, specifically natriuretic peptides (NPs) or C-reactive protein (CRP) enhance risk prediction (discrimination, calibration, reclassification) of AF when added to standard clinical factors.
  3. To examine whether a genotype score of top SNPs from CHARGE AF improves risk prediction (discrimination, calibration, reclassification) of AF over and above standard clinical factors.

Methods

  1. We have created a clinical risk prediction model in FHS. (Schnabel Lancet 2009; 373:739)
  2. We propose within each cohort to derive examine Cox proportional hazards function adjusting for age, sex, BMI, systolic blood pressure, hypertension treatment, PR interval history of heart murmur, and heart failure in relation to 5 and 10 year risk of incident atrial fibrillation (or flutter) using the FHS risk function. We may additionally adjust for race, and site (ARIC, CHS).
  3. We will examine test statistics including discrimination and calibration within each cohort.
  4. We may refine the FHS risk function depending on model fit.
  5. After we develop a risk prediction instrument that has good model fit in all cohorts we will examine the addition of biological and genetic markers.
  6. We will examine whether C-reactive protein and natriuretic peptides improve model calibration, discrimination and reclassification alone or together.
  7. We will examine whether a genotype score based on the results of CHARGE meta-analysis improves model calibration, discrimination and reclassification.

Follow-up of a SNP for Chronic Kidney Disease

PI: Caroline Fox, MD.MPH

Chronic kidney disease (CKD) affects 19 million adults in the United States, and is associated with cardiovascular disease, stroke, peripheral arterial disease, and all-cause mortality. Genetic factors play a role in the progression of renal disease. In the Framingham Heart Study, we have shown that kidney function is heritable, suggesting a role for genetic mechanisms in its etiology. Using genome-wide association, we have shown that a SNP in the UMOD gene is associated with CKD in the CHARGE consortium (in press, Nature Genetics). Because the results of this SNP showed robust inter-study results in all but the Framingham Heart Study (p=0.70), we obtained permission under the rapid approval mechanism (DNA proposal 2008.14) to genotype rs4293393. UMOD encodes uromodulin, the most abundant glycoprotein in the urine. Therefore, in a separate laboratory proposal, we designed a prospective nested case-control study to measure uromodulin levels in 100 individuals free of CKD who remain free of CKD after 8 years of follow-up, and 100 individuals free of CKD at baseline who develop CKD at follow-up. Our analysis of this laboratory study shows that cases have high uromodulin levels at baseline (p=0.02), even after adjusting for known CKD risk factors and baseline glomerular filtration rate.

We are now submitting a follow-up rapid DNA proposal to request permission to relate rs4293393 (already genotyped under 2008.14) to the urinary uromodulin levels that we have measured. The demonstration that a SNP in UMOD is associated with uromodulin levels, in concert with the results from the nested case-control laboratory study, would be proof-of-concept for uromodulin levels in the pathogenesis of CKD, and would demonstrate a functional consequence of rs4293393 (or a SNP in linkage disequilibrium). Therefore, we are requesting rapid approval from the DNA committee to relate this existing genotype data to existing uromodulin levels.

Follow-Up of on Meta-Analysis of Six Genome-wide Association Studies of MRI White Matter Hyper-intensities in the CHARGE Consortium.

PI: Sudha Seshadri, MD

We sought to identify genetic variants underlying cerebral white matter hyperintensities (WMH) by performing a prospective meta-analysis of genome-wide association data on the WMH measured on brain MRI in white subjects from six large cohort studies (the Framingham Heart Study, the Atherosclerosis Risk in Communities Study, the Cardiovascular Health Study, the Austrian Stroke Prevention Study, the Rotterdam Study, and the Aging Gene-Environment Susceptibility – Reykjavik Study) comprising the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) consortium. Each discovery cohort used genotype information to impute to HapMap’s CEU panel and age-, sex-, and total intracranial volume adjusted models to relate 2.2 million SNPs to WMH on brain MRI. At Framingham, we had 2320 participants from the Original and Offspring cohorts who had undergone brain MRI and had WMH measures available. After excluding 113 participants for presence of prevalent stroke (n=45), TIA (n=42), dementia (n=6) or other neurological conditions that might confound the diagnosis of covert MRI-infarcts on MRI (n=26), we had 2201 subjects (209 Original Cohort and 1992 Offspring participants) with WMH volume measures available.

Two genome-wide significant associations were found with rs7894407, an intronic SNP in the PDCD11 (Programmed cell death 11) gene (p= 2.7×10-8) and with rs2232267, an intronic SNP in C20orf103 (chromosome 20 open reading frame 1030 (p= 1.5×10-7). The SNP rs2232264 is located 341 bp away from rs2232267 and is in moderate LD with it (r2=0.43) in whites. It encodes a mis-sense mutation at amino-acid 103. This SNP has a p-value of 4×10-4 in the meta-analysis. Since it is a functional SNP we are asking to genotype it despite the high reliability of imputation.

Follow-up of GWAS results for all traits in the Framingham Heart Study

PI: Larry Atwood, PhD

In 2006 NHLBI funded the SHARe project that enabled GWA studies in Framingham. The SHARe project consisted of three steps. The first was to genotype the 550,000 SNPs in all the Framingham participants. The second was to make the data available to all qualified investigators and to perform association analysis of all traits with the 550K markers. The third step was to perform follow-up genotyping on the initial significant GWAS results (NHLBI provided resources to perform follow-up). The first two steps have now been completed. This request seeks permission to perform the follow-up genotyping, i.e. the third step.

Since the Affy 550K genotypes were made available (October 2007) we have made major progress in analysis. We have identified population substructure in order to control for it in GWAS. We have devoted significant resources to developing a software ‘pipeline’ that automates much of the tedious work in a GWAS. The investigator need only submit phenotypes to the pipeline which automatically performs association analysis on the 550K genotypes. In collaboration with the Broad Institute, we have also imputed the 550K genotypes to 2.5 million genotypes which have been incorporated into the pipeline for automated analysis.

In recognition that large sample sizes may be needed we have joined multiple consortium of studies who are pooling their results and performing meta-analysis. The primary consortium, that we took the lead in forming, is the Cohorts for Heart and Aging Research in Genetic Epidemiology (CHARGE); a group of studies that specialize in heart disease. To participate in CHARGE and other consortia we developed a meta analysis capability to combine the results from multiple studies. Framingham investigators have performed over 2000 GWAS in Framingham data and dozens of meta-analyses pooled across multiple studies. This has resulted in 29 publications that have been published, accepted, or submitted.

For this application each group of Framingham investigators have reviewed their results and constructed a list of their highest priority SNPs for follow-up genotyping. Our budget allows us to genotype 1536 SNPs on the Illumina platform. We pooled the list from all groups and use Illumina’s bioinformatic service to determine those SNPs with the highest probability of success using their technology. This is the list of SNPs that we seek permission to genotype. We will be using all three generations of the Framingham Heart Study. We request the GEN1, GEN2A, GEN2B, GEN3, and NOS plate sets. The total number of unique participants is 8116.

Association of SNPs Associated in Prior GWAS with Cardiovascular Disease Phenotypes

PI: Christipher O’Donnell, MD, MPH

Our group has an approved application to conduct research using SNPs from the Affymetrix 500K + Gene-focused 50K arrays and multiple phenotypes in the NHLBI’s Framingham SHARe database (dbGaP Application, PI O’Donnell). Framingham investigators have conducted over 2000 GWAS studies using ~2.5 million SNPs we imputed from HapMap in association with hundreds of phenotypes spanning a wide range of cardiovascular phenotypes including risk factors, biomarkers, subclinical vascular measures and clinically apparent cardiovascular disease phenotypes. In order to provide the strongest possible evidence for association, we have conducted in silico meta-analyses with GWAS data from multiple other cohorts collaborating in the Cohorts for Heart and Aging Research (CHARGE) Consortium. This effort has resulted in over 30 manuscripts or abstracts that have been submitted or accepted for publication. In addition, work from this effort has been presented in many presentations by our consortium scientists in international meetings. In this application, we propose to follow-up top SNPs that were not actually genotyped (or were poorly genotyped) on the Affymetrix 550K chips and were strongly associated with the following traits: blood pressure, liver function tests including bilirubin levels, platelet aggregation response, Raynaud’s syndrome, varicose veins, coronary artery calcium, hemostatic factor levels (fibrinogen, factor VII, factor VIII, von Willebrand Factor), leukocyte telomere length, estimated glomerular filtration rate, chronic kidney disease, visceral abdominal fat, body mass index, waist circumference, and uric acid. The rationale for genotyping these SNPs is to examine the association of the actual (not imputed) SNP with the primary phenotype from which the discovery arose, to enable participation in larger meta-analyses using the genotyped SNP, and to conduct additional secondary analyses of associations with other phenotypes that may be hypothesized to be associated with the SNP.

Additionally, over 1,300 SNPs have been reported to be associated with hundreds of diseases in 294 publications published GWAS reports to date. We propose to genotype a subset of SNPs within this set that have been reported in these published GWAS to be associated with cardiovascular diseases and traits that are available in the FHS dataset, including lipid levels (LDL, HDL and triglycerides), CRP, type 2 diabetes mellitus, cigarette abuse, obesity, myocardial infarction and related CVD outcomes, inflammatory diseases linked to CVD, and cardiovascular drug responsiveness. The rationale for genotyping these SNPs is to examine the association of the actual (not imputed) SNP with the primary phenotype from which the discovery arose, to enable participation in larger meta-analyses using the genotyped SNP, and to conduct additional secondary analyses of associations with other phenotypes that may be hypothesized to be associated with the SNP. Additionally, by genotyping these actual SNPs we will be able to conduct genetic risk score analyses using actually genotype SNPs.