Integrated associations of genotypes with multiple blood biomarkers linked to coronary heart disease risk - PubMed (original) (raw)

. 2009 Jun 15;18(12):2305-16.

doi: 10.1093/hmg/ddp159. Epub 2009 Mar 31.

Affiliations

Integrated associations of genotypes with multiple blood biomarkers linked to coronary heart disease risk

Fotios Drenos et al. Hum Mol Genet. 2009.

Abstract

Individuals at risk of coronary heart disease (CHD) show multiple correlations across blood biomarkers. Single nucleotide polymorphisms (SNPs) indexing biomarker differences could help distinguish causal from confounded associations because of their random allocation prior to disease. We examined the association of 948 SNPs in 122 candidate genes with 12 CHD-associated phenotypes in 2775 middle aged men (a genic scan). Of these, 140 SNPs indexed differences in HDL- and LDL-cholesterol, triglycerides, C-reactive protein, fibrinogen, factor VII, apolipoproteins AI and B, lipoprotein-associated phospholipase A2, homocysteine or folate, some with large effect sizes and highly significant P-values (e.g. 2.15 standard deviations at P = 9.2 x 10(-140) for F7 rs6046 and FVII levels). Top ranking SNPs were then tested for association with additional biomarkers correlated with the index phenotype (phenome scan). Several SNPs (e.g. in APOE, CETP, LPL, APOB and LDLR) influenced multiple phenotypes, while others (e.g. in F7, CRP and FBB) showed restricted association to the index marker. SNPs influencing six blood proteins were used to evaluate the nature of the associations between correlated blood proteins utilizing Mendelian randomization. Multiple SNPs were associated with CHD-related quantitative traits, with some associations restricted to a single marker and others exerting a wider genetic 'footprint'. SNPs indexing biomarkers provide new tools for investigating biological relationships and causal links with disease. Broader and deeper integrated analyses, linking genomic with transcriptomic, proteomic and metabolomic analysis, as well as clinical events could, in principle, better delineate CHD causing pathways amenable to treatment.

PubMed Disclaimer

Figures

Figure 1.

Figure 1.

Correlations between multiple phenotypes linked to CHD in 2775 men from the NPHSII study. Values in cells indicate Pearson's correlation coefficient R. *P < 0.01, **P < 0.001 (see colour code). Baseline and five repeat measures were available for cholesterol, triglycerides (TG), coagulation factor VII (FVIIc), fibrinogen, blood pressure (BP), smoking and body mass index (BMI) and single measures for the remaining traits.

Figure 2.

Figure 2.

Associations of 860 SNPs by chromosome with 12 blood phenotypes. The horizontal line indicates a critical FDR threshold of 0.2, approximately equivalent to a _P_-value < 10−3.

Figure 3.

Figure 3.

The relationship between minor allele frequency and effect size for SNP-phenotype associations exceeding the pre-specified FDR threshold. Effect size was expressed as: (A) variance and (B) as the standardized mean difference for comparisons of homozygous subjects. (C) Examples of SNPs with extreme effects assessed in terms of standardized mean difference.

Figure 4.

Figure 4.

Phenome plots. The gene of interest is depicted as an ellipse and the associated phenotypes traits as circles. The circle diameter is a measure of the variance of the phenotype explained by the variance of the genotypes (_R_2) of all the SNPs in the gene of interest. The numbers given next to each phenotype are the percent values of the coefficient of determination _R_2 of the phenotype for the combined effect of all the SNPs of the gene. The distance from the gene to each phenotype is a measure of the significance value, adjusted for multiple testing using the FDR, for the SNP with the strongest signal within the gene of interest. The number shown next to each edge is the percent value of the FDR adjusted _P_-value with its length measured from ellipse (gene) centre to circle (phenotype) centre. The dashed line represents those phenotypes which fall within the <0.1 FDR with the gene of interest. Those phenotypes and the gene are then expanded alongside in the accompanying figure so that details can be seen more clearly. For the phenome scan, an even more stringent FDR adjusted _P_-value of <0.1 to reduce further the risk of false-positive association. (Hcy, homocysteine; Flt, folate; SBP, systolic blood pressure; DBP, diastolic blood pressure; BMI, body mass index).

Figure 4.

Figure 4.

Phenome plots. The gene of interest is depicted as an ellipse and the associated phenotypes traits as circles. The circle diameter is a measure of the variance of the phenotype explained by the variance of the genotypes (_R_2) of all the SNPs in the gene of interest. The numbers given next to each phenotype are the percent values of the coefficient of determination _R_2 of the phenotype for the combined effect of all the SNPs of the gene. The distance from the gene to each phenotype is a measure of the significance value, adjusted for multiple testing using the FDR, for the SNP with the strongest signal within the gene of interest. The number shown next to each edge is the percent value of the FDR adjusted _P_-value with its length measured from ellipse (gene) centre to circle (phenotype) centre. The dashed line represents those phenotypes which fall within the <0.1 FDR with the gene of interest. Those phenotypes and the gene are then expanded alongside in the accompanying figure so that details can be seen more clearly. For the phenome scan, an even more stringent FDR adjusted _P_-value of <0.1 to reduce further the risk of false-positive association. (Hcy, homocysteine; Flt, folate; SBP, systolic blood pressure; DBP, diastolic blood pressure; BMI, body mass index).

Figure 5.

Figure 5.

The magnitude of the association between protein phenotypes is illustrated as the mean difference of the standardized levels of the phenotype between top and bottom tertile of the index phenotype, approximately equal to a 2 SD difference. When the most strongly associating SNP in the pQTGs is considered, the mean difference is between the two homozygote groups. (A) apoAI, (B) apoB, (C) CRP, (D) coagulation factor VII, (E) fibrinogen, and (F) Lp-PLA2.

Similar articles

Cited by

References

    1. Rader D.J., Daugherty A. Translating molecular discoveries into new therapies for atherosclerosis. Nature. 2008;451:904–913. - PubMed
    1. Hopkins P.N., Williams R.R. A survey of 246 suggested coronary risk factors. Atherosclerosis. 1981;40:1–52. - PubMed
    1. Brotman D.J., Walker E., Lauer M.S., O'Brien R.G. In search of fewer independent risk factors. Arch. Intern. Med. 2005;165:138–145. - PubMed
    1. Baigent C., Keech A., Kearney P.M., Blackwell L., Buck G., Pollicino C., Kirby A., Sourjina T., Peto R., Collins R., et al. Efficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from 90,056 participants in 14 randomised trials of statins. Lancet. 2005;366:1267–1278. - PubMed
    1. Willer C.J., Sanna S., Jackson A.U., Scuteri A., Bonnycastle L.L., Clarke R., Heath S.C., Timpson N.J., Najjar S.S., Stringham H.M., et al. Newly identified loci that influence lipid concentrations and risk of coronary artery disease. Nat. Genet. 2008;40:161–169. - PMC - PubMed

Publication types

MeSH terms

Substances

Grants and funding

LinkOut - more resources