Voxelwise gene-wide association study (vGeneWAS): multivariate gene-based association testing in 731 elderly subjects

Abstract

Imaging traits provide a powerful and biologically relevant substrate to examine the influence of genetics on the brain. Interest in genome-wide, brain-wide search for influential genetic variants is growing, but has mainly focused on univariate, SNP-based association tests. Moving to gene-based multivariate statistics, we can test the combined effect of multiple genetic variants in a single test statistic. Multivariate models can reduce the number of statistical tests in gene-wide or genome-wide scans and may discover gene effects undetectable with SNP-based methods. Here we present a gene-based method for associating the joint effect of single nucleotide polymorphisms (SNPs) in 18,044 genes across 31,662 voxels of the whole brain in 731 elderly subjects (mean age: 75.56±6.82SD years; 430 males) from the Alzheimer's Disease Neuroimaging Initiative (ADNI). Structural MRI scans were analyzed using tensor-based morphometry (TBM) to compute 3D maps of regional brain volume differences compared to an average template image based on healthy elderly subjects. Using the voxel-level volume difference values as the phenotype, we selected the most significantly associated gene (out of 18,044) at each voxel across the brain. No genes identified were significant after correction for multiple comparisons, but several known candidates were re-identified, as were other genes highly relevant to brain function. GAB2, which has been previously associated with late-onset AD, was identified as the top gene in this study, suggesting the validity of the approach. This multivariate, gene-based voxelwise association study offers a novel framework to detect genetic influences on the brain.

Figure 1

A histogram shows the minimum null P-values obtained from permutation tests. Data from 3 different voxels are shown on the same graph (blue, red, and black lines); they are obtained from 3 randomly chosen, uncorrelated voxels in the brain (5000 permutations each). The distributions are nearly identical, and agree with each other, as well as accurately reflecting the effective number of independent tests (M_eff).

Figure 2

Genetic association plots for univariate linear regression versus multi-locus PCReg. The -log₁₀(P-value) of each SNP in GRIN2B (a) and BEST3 (b) is plotted against its position in the gene. Each of the points is color coded by level of LD (compared to the top SNP, the purple diamond dot) as measured by r². The −log₁₀(P-value) of the gene-based PCReg test for each gene is overlaid on the plot for comparison (dotted black line). Plots were generated using the LocusZoom software (http://csg.sph.umich.edu/locuszoom/).

Figure 3

A color-coded significance map of the top gene at each voxel. Sections are shown at 8mm intervals throughout the brain. The top of each panel represents the anterior of the brain and bottom the posterior of the brain. The images are in radiological convention (the left side of the image is the patient’s right hemisphere). Color coding is based on the −log₁₀(P-value); warmer colors represent more significant associations.

Figure 4

Cluster sizes in vGeneWAS (red line) are compared with an average of simulated null maps (black line). We took the log₁₀ of the number of voxels in a cluster (not in mm³) across the brain in both maps for scaling purposes and for ease of comparison. The log₁₀ cluster sizes are then plotted using a density function such that the total area under each line is equal to 1. The average simulated null map contains a larger proportion of small cluster sizes than vGeneWAS (higher peaks in the black line at values close to the origin on the x-axis). The vGeneWAS map contains a larger proportion of large cluster sizes than the average simulated null map (the red line is higher at larger values and is more extended). A single slice view of the vGeneWAS and average simulated null cluster maps are pictured for comparison (inset). Every unique cluster is assigned its own color. There are more unique clusters than distinct colors making visual inspection difficult, but in general the clusters in the vGeneWAS maps are larger.

Figure 5

Regions in the brain associated with the top 5 genes from our vGeneWAS analysis (where the uncorrected P-value at a given voxel is overlaid on the minimum deformation template). The slices chosen best represent the regions where each gene was the most significantly associated gene in the brain. Images read from inferior to superior (left to right of the page) following radiological convention and with the top and bottom of each panel representing the anterior and posterior of the brain, respectively.

Figure 6

(a) The normalized histogram of observed P-values. The dashed line represents the cumulative distribution function (CDF) of Beta(1, 18044) where M_eff is based on the number of genes tested. The solid line represents the CDF of Beta(1, 15636) where M_eff is an estimate of the number of independent tests from permutation testing. (b) The Q-Q plot shows the observed P-values versus those expected from a Beta(1, 15636) distribution (black dots). The solid gray line represents a purely null distribution of P-values. (c) The histogram of corrected P-values (P_c) approximately follows a uniform distribution. (d) The Q-Q plot of the observed P_c versus those expected from a null distribution.

Figure 7

Map of P-values for GAB2 at every voxel in the brain after correction for multiple comparisons across voxels (but not corrected for search across the genome, as we are only testing one gene) using the original FDR method. P-values significant after FDR correction (at q = 3.4×10⁻⁴) are color-coded. Warmer colors are more significant. GAB2 has a more distributed effect on the brain than is evident in the vGeneWAS results.

Figure 8

Genetic association plot, for different SNPs in the GAB2 gene, at the top voxel from our analysis. The -log₁₀(P-value) of each SNP in GAB2 is plotted against its position in the gene. Each of the points is color coded by level of LD (compared to the top SNP, the purple diamond dot) as measured by r². The −log₁₀(P-value) of the gene-based PCReg test for GAB2 at this voxel is overlaid on the plot for comparison (dotted black line). In this case, the gene-based test shows a greater effect size than univariate tests on any of the component SNPs treated independently. This shows that the gene-based test can be more powerful than performing separate tests on component SNPs.

Figure 9

vGeneWAS may control the false discovery rate better than vGWAS. The cumulative distribution function (CDF) of P_c-values from vGeneWAS (solid green line) is compared to the CDF of P_c-values from vGWAS (Stein et al. 2010a; solid black line). Three lines represent different correction thresholds of q = 0.05 (red dashed line), q = 0.30 (black dashed line), and q = 0.50 (blue dotted line).