Background:

There are now many powerful GWAS based on large sets of meta-analysed data from multiple studies and the number of GWAS in literature is rapidly increasing. These studies provide robust effect estimates for the associations of single nucleotide polymorphisms (SNPs) across the genome with phenotypes of interest. For example, associations for all SNPs across the genome in relation to anthropometric phenotypes are available from the GIANT (Genetic Investigation of ANthropometric Traits) consortium involving meta-analysed data on up to 250,000 individuals (1-3). The GIANT GWAS observed many loci predicting variation in BMI (32 loci) and fat distribution (29 loci) at genome-wide significance level.

However, many loci will exist beyond those established using the stringent genome-wide cut-off level that are in reality reliably related to the phenotype of interest. Data from the GIANT height GWAS (N=183,727 individuals) identified 180 loci associated with height at genome-wide significance levels, explaining 10% of the phenotypic variation in height. However, the authors further estimated, based on recently developed methods (4), that there were in fact 697 loci (95% CI: 483-1040) with effects greater or equal to those identified from the GWAS which would explain a total of 15.7% of the phenotypic variation in height. Preliminary data from the GIANT consortium supports the conclusion that many of the associations that were observed but did not reach genome-wide significance are in fact real associations (J. Hirschhorn, Personal Communication).

Furthermore, when all SNPs across the genome are fitted simultaneously and the variance explained by all the SNPs together is estimated, 45% of the phenotypic variance in height can be explained (5).Together with the fact that for SNPs explaining 0.01% variation there is 99% power to get the direction of the association correct in a sample size of 129,000 individuals, this suggests that robust GWAS with large sample sizes can reliably inform the development of genome-wide polygenic instruments.

The inclusion of a potentially vast number of 'risk-alleles' across the genome would therefore explain significantly more phenotypic variation than instruments based on individual genetic variants or several established genetic variants together. As such, genome-wide polygenic scores provide a platform from which to develop a novel application of the Mendelian Randomization approach using genome-wide instruments that would incorporate a much larger proportion of genetic information than has otherwise been explored previously, resulting in considerably greater power for detecting causal effects for novel phenotypes of interest.

Aims:

The aims of this project are to develop and apply an original genome-wide Mendelian Randomization approach in order to: i) identify of novel causal risk factors for adult CVD, and ii) examine their impact on childhood and adolescent CVD-related traits

Hypotheses:

We hypothesise that the application of genome-wide allele scores for indexing traits known to be related to CVD-related phenotypes, such as adiposity, lipids and inflammation, will accurately index CVD risk and that these scores can subsequently be extended to explore more novel predictors such as iron status, fatty acid profile and uric acid.

Exposure variables:

genome-wide SNP data and phenotypic data for established CVD-related phenotypes (BMI, lipids, inflammatory markers) and novel predictors in CVD (iron status makers, fatty acids, uric acid)

Outcome variables:

Blood pressure, flow-mediated dilation, intima-media thickness

Confounders:

Maternal eduation, paternal education, family income, parental occupational class, child IQ, maternal smoking

References:

1. Speliotes EK, Willer CJ, Berndt SI, et al. Association analyses of 249,796 individuals reveal 18 new loci associated with body mass index. Nat Genet 2010; 42(11): 937-48

2. Lango Allen H, Estrada K, Lettre G, et al. Hundreds of variants clustered in genomic loci and biological pathways affect human height. Nature 2010; 467(7317): 832-8

3.Heid IM, Jackson AU, Randall JC, et al. Meta-analysis identifies 13 new loci associated with waist-hip ratio and reveals sexual dimorphism in the genetic basis of fat distribution. Nat Genet 2010; 42(11): 949-60

4. Park JH, Wacholder S, Gail MH, et al. Estimation of effect size distribution from genome-wide association studies and implications for future discoveries. Nat Genet 2010; 42(7): 570-5

5.Yang J, Benyamin B, McEvoy BP, et al. Common SNPs explain a large proportion of the heritability for human height. Nat Genet 2010; 42(7): 565-9