JACC
HOME SUBSCRIPTIONS CURRENT ISSUE PAST ISSUES CARDIOSOURCE SEARCH HELP FEEDBACK
QUICK SEARCH:   [advanced]


     

J Am Coll Cardiol, 2006; 48:10-14, doi:10.1016/j.jacc.2006.06.046 (Published online 16 October 2006).
© 2006 by the American College of Cardiology Foundation
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
j.jacc.2006.06.046v1
48/9_Suppl_A/A10    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Cohen, J. C.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Cohen, J. C.

STATE-OF-THE-ART PAPER

Genetic Approaches to Coronary Heart Disease

Jonathan C. Cohen, PhD*

Center for Human Nutrition and McDermott Center for Human Growth and Development, University of Texas Southwestern Medical Center at Dallas, Dallas, Texas.

Manuscript received November 29, 2005; revised manuscript received May 22, 2006, accepted June 20, 2006.

* Reprint requests and correspondence: Dr. Jonathan C. Cohen, Center for Human Nutrition, UT Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, Texas 75390. (Email: jonathan.cohen{at}utsouthwestern.edu).


   Abstract
Top
 Abstract
References
 
Developments in high-throughput genotyping have progressed to the point where genome-wide association studies are becoming practical. Multistage designs involving large numbers of sequence variants (~300,000) and relatively large samples sizes (several hundred cases and control subjects) will be essential to reliably detect alleles with appreciable effect sizes (2-fold increase in relative risk). Direct sequencing of candidate genes in cases and control subjects provides an alternative approach that can reveal low-frequency alleles that influence disease susceptibility. Ultimately, the outcome of both approaches will depend on the genetic architecture of coronary heart disease.

Abbreviations and Acronyms
  CHD = coronary heart disease
  LDL-C = low-density lipoprotein cholesterol
  SNP = single nucleotide polymorphism

Although the past 2 decades have brought considerable advances in the detection and treatment of coronary heart disease (CHD), CHD remains a leading cause of death and disability in Western countries (1). The primary cause of CHD is atherosclerosis, a progressive chronic disease process characterized by the accumulation of lipids and fibrous elements in medium-sized and large arteries. Epidemiologic studies have revealed numerous risk factors that influence the development and severity of atherosclerosis (2). Whereas some major risk factors for atherosclerosis, such as smoking, are largely environmental, family and twin studies have consistently found that genetic factors contribute importantly to premature CHD (3). Among identical twins, who share all of their genes in common, premature cardiac death confers an 8-fold increase in risk to the surviving male sibling and a 15-fold increase in risk to female siblings (4). The corresponding risk is significantly lower in fraternal twins, who share half of their genes in common (2.6 for women and 3.8 for men). Family history is a significant risk factor for atherosclerosis, and the contribution of family history cannot be fully accounted for by known cardiac risk factors (4–6). Genetic factors also contribute significantly to most of the major risk factors for CHD (diabetes, hypertension, elevated plasma levels of low-density lipoprotein cholesterol [LDL-C], and low levels of plasma high-density lipoprotein cholesterol), and yet the contribution of family history is not fully explained by known cardiac risk factors, suggesting that other yet-to-be-determined genetic factors also contribute to cardiovascular risk. Identification of these factors could provide a molecular handle on new pathways that contribute to the susceptibility to atherosclerosis and provide new therapeutic targets for the prevention and treatment of CHD. In addition, genetic testing for sequence variations associated with coronary atherosclerosis may assist in the risk stratification of asymptomatic individuals. For example, if the subset of asymptomatic individuals who are likely to develop heart disease can be more accurately identified by analysis of single or multiple single nucleotide polymorphisms (SNPs), interventions, such as the institution of aggressive lipid-lowering therapy, could be initiated sooner to prevent death and disability from coronary atherosclerosis (7).

The identification of sequence variants that confer susceptibility to CHD has been hampered by 2 major obstacles. First, the pathogenesis of CHD is extremely complex, involving interactions between various cells in the vessel wall, circulating lipoproteins, inflammatory mediators, and the clotting and fibrinolytic cascades. Therefore, the effect of any single sequence variant on susceptibility is likely to be small. Second, the human genome is highly polymorphic (8). Any 2 homologous chromosomes contain an average of 1 sequence difference per ~1,000 nucleotides, and the total reservoir of sequence variation in the population is even greater; resequencing studies have revealed that most genes contain dozens of SNPs. Because most polymorphisms probably have little or no effect on gene function, this enormous genetic heterogeneity has confounded efforts to identify individual SNPs that systematically influence phenotypes.

The complex pathogenesis of atherosclerosis poses a significant problem for approaches based on genetic linkage, which have limited power to detect loci with modest effects (9). Several groups have undertaken genome-wide linkage scans of cardiovascular disease (10–14). The results of these scans have been inconclusive; Pajukanta et al. (14) reported that loci on chromosome 2q21-22 and Xq23-26 were linked to premature CHD in 156 Finnish families; Francke et al. (11) reported linkage to 3 regions on chromosomes 3q27, 8q23, and 16p13-pter in 99 independent families of northeast Indian origin; Harrap et al. (10) found that a locus on chromosome 2q36-37 was associated with acute coronary syndrome in 61 sibling pairs aged less than 70; Broeckel et al. (12) identified a single chromosomal region on chromosome 14 that was linked to myocardial infarction in 513 families; Hauser et al. (13) identified regions on chromosomes 3q13 and 5q31 that were linked to premature CHD (<51 years in men and <56 years in women) in 493 affected sibling pairs. Wang et al. (15) reported a region on chromosome 1p34-36 that was associated with premature CHD in 428 multiplex families. Therefore, each of these linkage studies has identified different loci, but none has proved reproducible in different cohorts or led to the identification of genes that are systematically associated with CHD susceptibility. Chiodini et al. (16) performed a meta-analysis of 4 published genome-wide screens and concluded that the strongest evidence for linkage was for loci on chromosomes 3q26-27 and 2q34-37, but the intervals identified were very large (30 cM), and the specific genes and sequence variants responsible for these linkages have not been identified.

The potential complications of using linkage analysis to identify genes responsible for atherosclerosis is further illustrated by a report from Wang et al. (17), who used linkage to map the defect responsible for severe premature atherosclerosis that appeared to segregate as an autosomal-dominant disease in an extended pedigree. Sequencing of candidate genes in the linked region revealed a 21 base pair (bp) deletion in MEF2A, a gene expressed in the coronary vasculature that was previously implicated in cardiac development. The deletion was not found in 119 individuals with normal angiograms (17). The authors showed that the mutant protein acted as a dominant negative in tissue culture studies and proposed that the associated increase in genetic risk was due to defects in endothelial function. These same authors subsequently identified several other missense mutations in MEF2A and concluded that 2% of cases of premature atherosclerosis are caused by mutations in this gene (18). More recently, however, Weng et al. (19) used direct DNA sequencing to screen MEF2A in 300 individuals with premature coronary artery disease and in 300 healthy elderly control subjects. Only a single missense mutation (S360L) was identified, and the affected residue was not evolutionarily conserved. Moreover, 3 of the control individuals had the 21 bp deletion reported by Wang et al. (17). Analysis of the extended families of the 3 probands revealed no correlation between the presence of the mutation and CHD.

The difficulties encountered in genetic linkage studies have stimulated renewed interest in genetic association studies, which have been used extensively to examine the relationship between sequence variations in candidate genes and phenotypes related to CHD, including plasma lipid levels, blood pressure, and diabetes, and end points such as myocardial infarction, carotid intima-media thickness, and coronary artery calcification (3). Genetic association provides more power to detect modest allelic effects (9). However, the highly polymorphic nature of the genome makes it extremely unlikely that an arbitrarily selected candidate SNP will prove to be functionally significant. Of the literally thousands of association studies reported, the overwhelming majority considered a single SNP, or at most a handful of SNPs, in a single candidate gene. Whereas many significant associations have been reported, very few have proved consistent in different studies. Thus, the specific DNA sequence variations that confer susceptibility to CHD remain elusive.

Recently, it has become apparent that the power of association studies can be applied on a genome-wide scale. In a landmark paper, Risch and Merikangas (20) proposed that loci with modest phenotypic effects could be detected by genome-wide association studies. Analysis of a dense panel of SNPs spanning the genome would obviate the need for a priori selection of candidate genes, the major limitation of traditional association studies. Advances in genotyping technologies, coupled with decreasing genotyping costs and the availability of high-density SNP maps across the genome, have brought analytical strategies based on large-scale allelic association within reach (21). Several investigators have examined the requirements for successful identification of disease-causing sequence variants using this approach. These considerations have centered around 4 crucial issues:

1 How many SNPs are needed for a whole-genome scan? A whole-genome association scan is based on the concept of indirect association: the SNPs assayed in the scan serve as markers for SNPs that are not assayed directly (9). Association studies are performed on unrelated individuals; therefore, the region of identity around the marker SNPs is much smaller than is the case in linkage studies. Accordingly, the number of marker SNPs needed to provide an informative scan of the genome is much higher for association studies than for linkage analyses. The number of SNPs required to provide comprehensive coverage of the genome is a subject of ongoing debate. Simulation studies have yielded a wide range of estimates, from as few as 30,000 to 50,000 (22) to as many as 1,000,000 (23). The most direct analysis of this question was undertaken by Ke et al. (24), who analyzed a dense set of 5,000 markers spanning 10 Mb on chromosome 20 in European Caucasians, Centre d’Etude du Polymorphisme Humain Caucasians, and African Americans. Extrapolation from this data set to the entire genome indicated that genome-wide coverage of regions in strong-linkage disequilibrium would require 100,000 to 300,000 SNPs in Western Europeans. More recently, Hinds et al. (25) characterized patterns of DNA sequence variation across the genome by genotyping 1,586,383 (SNPs) in 71 Americans of European, African, and Asian ancestry. Comparison of this data with direct DNA sequencing of 152 genes from the same individuals indicated that a set of tagging SNPs comprising 8% of all SNPs in the regions surveyed ascertained 70% of all common sequence variants in the regions (i.e., those with a minor allele frequency of >0.1) with an r2 value of >0.8. Extrapolating this finding to the whole genome indicated that an efficient genome-wide scan could be performed with ~300,000 SNPs.
2 How can true associations be distinguished from false positives when tens of thousands of statistical tests are required? Analysis of the very large number of SNPs required for a genome-wide association study presents the classical statistical problem of multiple testing with the resulting inflation of the overall type I error rate (i.e., the number of false-positive tests) due to the large number of hypothesis tests being evaluated (26). The type I error rates are of 2 types: the pointwise error rate (nominal significance level, alpha), which is the probability of erroneously rejecting a null hypothesis when making a single test, and the familywise error rate, which is the probability of erroneously rejecting at least 1 null hypothesis when making k (>1) tests using data from a single study (27). The familywise error rate approaches 100% very quickly as the number of tests increases. For example, with only 5 independent tests at a pointwise error rate of 0.05 (alpha), the familywise error rate is 23%; with 100 tests, it is 99.4%. By adjusting the pointwise error rate, the familywise error rate can be controlled.
Two main statistical approaches to accommodate multiple testing have been developed, one based on the familywise error rate, and another based on the false discovery rate (26). Familywise error rates can be accommodated by simply adjusting the nominal significance level for the number of tests performed, for example by using the Bonferroni correction. This approach ensures that the familywise type I error rate is maintained. For example, if 100 polymorphic markers are considered and the nominal significance level of 0.05 is divided by the number of tests performed, a nominal level of 0.0005 will lead to a 0.05 familywise error rate. Because most SNPs are likely to have only modest effects on coronary atherosclerosis, even true associations are unlikely to achieve high p values except in very large samples. Therefore, the application of strategies that effectively increase the stringency required for a significant p value, such as the Bonferroni correction, are likely to result in an unacceptably high rate of type II errors. To achieve a better balance between type I and type II error rates, approaches have been developed based on the false discovery rate. The false discovery rate is the expected proportion of errors among the rejected hypotheses. Rather than controlling the chance of any false positives (as the Bonferroni correction does), the false discovery rate controls the expected proportion of false positive tests based on the observed distribution of p values. Several approaches based on the false discovery rate have been developed (28), but the performance of these methods when applied to genome-wide association data has not been systematically evaluated.
The high rate of false-positive associations in published reports has lead several authors to propose study designs based on replication, rather than statistical correction, to limit type I error rates in genome-wide association studies (29–32). Each of these designs incorporates a screening stage, in which a large number of markers are screened at a low stringency, and a replication stage, in which the most promising markers are screened in a second population (Fig. 1). Extensive simulations under several different study configurations have shown that the staged method is more cost effective and more powerful than conventional case-control study designs (29,32).
3 How can false-positive associations that are due to population substructure be prevented? An important prerequisite of genetic association studies is that the cases and control subjects have the same genetic ancestry. Differences in genetic ancestry (i.e., population substructure) can give rise to true case-control differences in allele frequency that result in false-positive genetic associations (33). Undetected population substructure is a potential problem for every association study (34). Fortuitously, the level of population structure in many ethnic groups is low, and it has been suggested by some that the problem of substructure is limited if gross levels of population structure are avoided (35). Others have shown that problems associated with population substructure increase markedly with increasing sample size (34). In large samples, modest levels of population structure can lead to false-positive associations, thus presenting an intrinsic conundrum: large sample sizes are essential to detect the modest effects of sequence variations contributing to complex traits, yet false-positive rates increase in proportion to sample size. Several statistical methods have been developed that can be used to assay for population substructure in large samples. These tests can be classified into 2 categories: those based on "genomic control" and those based on "structured association" (36). Genomic control methods are based on the assumption that the effect of population structure is to distort the null distribution of the test statistic by a factor {lambda}. In this case {lambda} can be estimated empirically from the marker loci and used to rescale the test statistic to match theoretical expectation (37). Structured association methods infer the details of population structure en route to testing for association. The association analysis can then be conditioned on the inferred ancestries of individuals by considering associations within subpopulations (36). Alternatively, DNA markers can be used to match cases and control subjects. Hinds et al. (38) showed that relatively simple matching strategies based on a set of 300 anonymous DNA markers can effectively mitigate the impact of population stratification.
4 What sample sizes are required to detect true associations? The sample sizes required to detect genetic associations are determined by the underlying architecture of the trait: the frequency of the susceptibility allele and the size of its effect on the trait. Wang et al. (39) pointed out that, for low-frequency alleles with small phenotypic effects, unrealistically large sample sizes will be required. For example, detection of susceptibility alleles with frequencies below 0.1 and effect sizes less than an odds ratio of 1.3 will require more than 10,000 cases and 10,000 control subjects to achieve convincing statistical support for a disease association. The proportion of disease-susceptibility alleles that lie outside this range and will, therefore, be tractable to genome-wide association approaches is not known. Accordingly, the question of power is a major consideration for genome-wide association studies. Calculations by Thomas et al. (40) indicate that relatively large sample sizes will be required for reliable detection even for common alleles with relatively large effects. For example, detection of an allele that is present in 20% of the population and that doubles the risk of disease will require a total of ~600 cases and 600 control subjects in a 2-stage design.


Figure 1
View larger version (30K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1 Multistage design for genome-wide association study. A multistage design with replication allows testing at a nominal significance threshold with elimination of false-positive associations by replication. Validation in an independent population is essential, with a larger sample size at this step to protect against false-negative associations. SNP = single nucleotide polymorphism.

 
An alternative to genomewide screens based on linkage disequilibrium is to select sequence variants that are most likely to affect gene function. Rather than selecting candidate genes, a large number of potentially functional variants across the genome can be assayed. Recently, Shiffman et al. (41) reported the results of a genome-wide screen for association of functional variants and coronary atherosclerosis. In that study, analysis of 11,053 SNPs in 340 cases and 346 control subjects revealed 637 that were associated with myocardial infarction at a nominal significance threshold of 0.05. All 637 were then tested in a second set of cases and control subjects, and 31 achieved the nominal significance threshold and had the same risk allele as the first group. These 31 sequence variants were then tested in the third group, and 7 achieved a nominal 1-sided significance threshold of 0.05, of which 4 were significant using a false discovery rate of 10%. These 4 variants were associated with genes encoding the cytoskeletal protein palladin, a tyrosine kinase (ROS1), and 2 G protein-coupled receptors (TAS2R50 and OR13G1). Under a simple probability model, none of the 11,053 sequence variants in the initial analysis would be expected to achieve the nominal significance threshold in all 3 case-control groups simply by chance. Further studies will now be required to determine whether these sequence variants are systematically associated with CHD.

Variation in disease susceptibility may also arise from the cumulative effects of multiple sequence variants, each of which is present at low frequency in the population. Association studies have little power to detect the effects of such variants (40). Our laboratory has shown that low-frequency alleles contribute to plasma levels of lipoproteins (42–44), and has used these variants to examine the relationship between lifelong reduction in plasma levels of LDL-C and CHD. Through direct sequencing of candidate genes in individuals with low (below the 5th percentile) or high (above the 95th percentile) plasma levels of LDL-C, we identified 3 sequence variants in the gene encoding PCSK9 that lower plasma levels of LDL-C (43,45). Two of these mutations completely abolish PCSK9 protein function and are present in ~2.5% of African Americans. These mutations are associated with a 30% to 40% reduction in plasma levels of LDL-C and a marked reduction (88%) in CHD (46). The third mutation is less severe and is found in ~3% of whites. This mutation lowers LDL-C levels by 15% to 20% and decreases incident CHD by 50%.

Whereas these low-frequency alleles explain only a small fraction of the CHD burden in the population, they have important clinical implications. First, the results suggest that PCSK9 is an attractive target for therapeutic intervention. Heterozygotes for PCSK9 deficiency apparently have no adverse clinical sequelae. Therefore, specific pharmacologic inhibition of PCSK9 could potentially reduce LDL-C levels and CHD with few side effects.

Second, the reduction in CHD associated with PCSK9 mutations is greater than would be predicted from cholesterol-lowering trials (46). This finding suggests that earlier intervention with cholesterol-lowering drugs may compound the beneficial effects of these agents on CHD.


   Footnotes
 
This work was supported by a grant from the Leducq Foundation.


   References
Top
 Abstract
 References
 

  1. Grundy SM, Cleeman JI, Merz CN, et al. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines J Am Coll Cardiol 2004;44:720-732.[Abstract/Free Full Text]
  2. Adult Treatment Panel III Third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report Circulation 2002;106:3143-3421.[Medline]
  3. Lusis AJ, Mar R, Pajukanta P. Genetics of atherosclerosis Annu Rev Genomics Hum Genet 2004;5:189-218.[CrossRef][ISI][Medline]
  4. Marenberg ME, Risch N, Berkman LF, Floderus B, de Faire U. Genetic susceptibility to death from coronary heart disease in a study of twins N Engl J Med 1994;330:1041-1046.[Abstract/Free Full Text]
  5. Nasir K, Michos ED, Rumberger JA, et al. Coronary artery calcification and family history of premature coronary heart disease: sibling history is more strongly associated than parental history Circulation 2004;110:2150-2156.[CrossRef][ISI][Medline]
  6. Snowden CB, McNamara PM, Garrison RJ, Feinleib M, Kannel WB, Epstein FH. Predicting coronary heart disease in siblings—a multivariate assessment: the Framingham Heart Study Am J Epidemiol 1982;115:217-222.[Abstract/Free Full Text]
  7. Heart Protection Study Collaborative Group MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial Lancet 2002;360:7-22.[CrossRef][ISI][Medline]
  8. Livingston RJ, von Niederhausern A, Jegga AG, et al. Pattern of sequence variation across 213 environmental response genes Genome Res 2004;14:1821-1831.[Abstract/Free Full Text]
  9. Carlson CS, Eberle MA, Kruglyak L, Nickerson DA. Mapping complex disease loci in whole-genome association studies Nature 2004;429:446-452.[CrossRef][Medline]
  10. Harrap SB, Zammit KS, Wong ZY, et al. Genome-wide linkage analysis of the acute coronary syndrome suggests a locus on chromosome 2 Arterioscler Thromb Vasc Biol 2002;22:874-878.[Abstract/Free Full Text]
  11. Francke S, Manraj M, Lacquemant C, et al. A genome-wide scan for coronary heart disease suggests in Indo-Mauritanians a susceptibility locus on chromosome 16p13 and replicates linkage with the metabolic syndrome on 3q27 Hum Mol Genet 2001;10:2751-2765.[Abstract/Free Full Text]
  12. Broeckel U, Hengstenberg C, Mayer B, et al. A comprehensive linkage analysis for myocardial infarction and its related risk factors Nat Genet 2002;30:210-214.[CrossRef][ISI][Medline]
  13. Hauser ER, Crossman DC, Granger CB, et al. A genomewide scan for early-onset coronary artery disease in 438 families: the GENECARD Study Am J Hum Genet 2004;75:436-447.[CrossRef][ISI][Medline]
  14. Pajukanta P, Cargill M, Viitanen L, et al. Two loci on chromosomes 2 and X for premature coronary heart disease identified in early- and late-settlement populations of Finland Am J Hum Genet 2000;67:1481-1493.[CrossRef][ISI][Medline]
  15. Wang Q, Rao S, Shen GQ, et al. Premature myocardial infarction novel susceptibility locus on chromosome 1P34-36 identified by genomewide linkage analysis Am J Hum Genet 2004;74:262-271.[CrossRef][ISI][Medline]
  16. Chiodini BD, Lewis CM. Meta-analysis of 4 coronary heart disease genome-wide linkage studies confirms a susceptibility locus on chromosome 3q Arterioscler Thromb Vasc Biol 2003;23:1863-1868.[Abstract/Free Full Text]
  17. Wang L, Fan C, Topol SE, Topol EJ, Wang Q. Mutation of MEF2A in an inherited disorder with features of coronary artery disease Science 2003;302:1578-1581.[Abstract/Free Full Text]
  18. Bhagavatula MR, Fan C, Shen GQ, et al. Transcription factor MEF2A mutations in patients with coronary artery disease Hum Mol Genet 2004;13:3181-3188.[Abstract/Free Full Text]
  19. Weng L, Kavaslar N, Ustaszewska A, et al. Lack of MEF2A mutations in coronary artery disease J Clin Invest 2005;115:1016-1020.[CrossRef][ISI][Medline]
  20. Risch N, Merikangas K. The future of genetic studies of complex human diseases Science 1996;273:1516-1517.[ISI][Medline]
  21. Heller MJ. DNA microarray technology: devices, systems, and applications Annu Rev Biomed Eng 2002;4:129-153.[CrossRef][ISI][Medline]
  22. Taillon-Miller P, Saccone SF, Saccone NL, et al. Linkage disequilibrium maps constructed with common SNPs are useful for first-pass disease association screens Genomics 2004;84:899-912.[CrossRef][ISI][Medline]
  23. Gabriel SB, Schaffner SF, Nguyen H, et al. The structure of haplotype blocks in the human genome Science 2002;296:2225-2229.[Abstract/Free Full Text]
  24. Ke X, Hunt S, Tapper W, et al. The impact of SNP density on fine-scale patterns of linkage disequilibrium Hum Mol Genet 2004;13:577-588.[Abstract/Free Full Text]
  25. Hinds DA, Stuve LL, Nilsen GB, et al. Whole-genome patterns of common DNA variation in three human populations Science 2005;307:1072-1079.[Abstract/Free Full Text]
  26. Curran-Everett D. Multiple comparisons: philosophies and illustrations Am J Physiol Regul Integr Comp Physiol 2000;279:R1-R8.[Abstract/Free Full Text]
  27. Weller JI, Song JZ, Heyen DW, Lewin HA, Ron M. A new approach to the problem of multiple comparisons in the genetic dissection of complex traits Genetics 1998;150:1699-1706.[Abstract/Free Full Text]
  28. Sabatti C, Service S, Freimer N. False discovery rate in linkage and association genome screens for complex disorders Genetics 2003;164:829-833.[Abstract/Free Full Text]
  29. Satagopan JM, Elston RC. Optimal two-stage genotyping in population-based association studies Genet Epidemiol 2003;25:149-157.[CrossRef][ISI][Medline]
  30. Satagopan JM, Venkatraman ES, Begg CB. Two-stage designs for gene-disease association studies with sample size constraints Biometrics 2004;60:589-597.[CrossRef][ISI][Medline]
  31. Satagopan JM, Verbel DA, Venkatraman ES, Offit KE, Begg CB. Two-stage designs for gene-disease association studies Biometrics 2002;58:163-170.[CrossRef][ISI][Medline]
  32. Saito A, Kamatani N. Strategies for genome-wide association studies: optimization of study designs by the stepwise focusing method J Hum Genet 2002;47:360-365.[CrossRef][ISI][Medline]
  33. Goldstein DB, Chikhi L. Human migrations and population structure: what we know and why it matters Annu Rev Genomics Hum Genet 2002;3:129-152.[CrossRef][ISI]
  34. Marchini J, Cardon LR, Phillips MS, Donnelly P. The effects of human population structure on large genetic association studies Nat Genet 2004;36:512-517.[CrossRef][ISI][Medline]
  35. Wacholder S, Rothman N, Caporaso N. Counterpoint: bias from population stratification is not a major threat to the validity of conclusions from epidemiological studies of common polymorphisms and cancer Cancer Epidemiol Biomarkers Prev 2002;11:513-520.[Free Full Text]
  36. Pritchard JK, Donnelly P. Case-control studies of association in structured or admixed populations Theor Popul Biol 2001;60:227-237.[CrossRef][ISI][Medline]
  37. Devlin B, Roeder K. Genomic control for association studies Biometrics 1999;55:997-1004.[CrossRef][ISI][Medline]
  38. Hinds DA, Stokowski RP, Patil N, et al. Matching strategies for genetic association studies in structured populations Am J Hum Genet 2004;74:317-325.[CrossRef][ISI][Medline]
  39. Wang WY, Barratt BJ, Clayton DG, Todd JA. Genome-wide association studies: theoretical and practical concerns Nat Rev Genet 2005;6:109-118.[CrossRef][ISI][Medline]
  40. Thomas DC, Haile RW, Duggan D. Recent developments in genomewide association scans: a workshop summary and review Am J Hum Genet 2005;77:337-345.[CrossRef][ISI][Medline]
  41. Shiffman D, Ellis SG, Rowland CM, et al. Identification of four gene variants associated with myocardial infarction Am J Hum Genet 2005;77:596-605.[CrossRef][ISI][Medline]
  42. Cohen JC, Kiss RS, Pertsemlidis A, Marcel YL, McPherson R, Hobbs HH. Multiple rare alleles contribute to low plasma levels of HDL cholesterol Science 2004;305:869-872.[Abstract/Free Full Text]
  43. Cohen J, Pertsemlidis A, Kotowski IK, Graham R, Garcia CK, Hobbs HH. Low LDL cholesterol in individuals of African descent resulting from frequent nonsense mutations in PCSK9 Nat Genet 2005;37:161-165.[CrossRef][ISI][Medline]
  44. Cohen J, Pertsemlidis A, Fahmi S, et al. Multiple rare variants in NPC1L1 associated with reduced sterol absorption and plasma low-density lipoprotein levels Proceed Natl Acad Sci U S A 2006;103:1810-1815.
  45. Kotowski IK, Pertsemlidis A, Luke A, et al. A spectrum of PCSK9 alleles contributes to plasma levels of low-density lipoprotein cholesterol Am J of Hum Genet 2006;78:410-422.
  46. Cohen JC, Boerwinkle E, Mosley Jr. TH, Hobbs HH. Sequence variations in PCSK9, low LDL, and protection against coronary heart disease N Engl J Med 2006;354:34-42.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 J Am Coll CardiolHome page
O. A. Iakoubova, C. H. Tong, C. M. Rowland, T. G. Kirchgessner, B. A. Young, A. R. Arellano, D. Shiffman, M. S. Sabatine, H. Campos, C. J. Packard, et al.
Association of the Trp719Arg polymorphism in kinesin-like protein 6 with myocardial infarction and coronary heart disease in 2 prospective trials: the CARE and WOSCOPS trials.
J. Am. Coll. Cardiol., January 29, 2008; 51(4): 435 - 443.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
j.jacc.2006.06.046v1
48/9_Suppl_A/A10    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Cohen, J. C.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Cohen, J. C.

HOME SUBSCRIPTIONS CURRENT ISSUE PAST ISSUES CARDIOSOURCE SEARCH HELP FEEDBACK