Discovering genetic associations with high-dimensional neuroimaging phenotypes: A sparse reduced-rank regression approach

Abstract

There is growing interest in performing genome-wide searches for associations between genetic variants and brain imaging phenotypes. While much work has focused on single scalar valued summaries of brain phenotype, accounting for the richness of imaging data requires a brain-wide, genome-wide search. In particular, the standard approach based on mass-univariate linear modelling (MULM) does not account for the structured patterns of correlations present in each domain. In this work, we propose sparse Reduced Rank Regression (sRRR), a strategy for multivariate modelling of high-dimensional imaging responses (measurements taken over regions of interest or individual voxels) and genetic covariates (single nucleotide polymorphisms or copy number variations) that enforces sparsity in the regression coefficients. Such sparsity constraints ensure that the model performs simultaneous genotype and phenotype selection. Using simulation procedures that accurately reflect realistic human genetic variation and imaging correlations, we present detailed evaluations of the sRRR method in comparison with the more traditional MULM approach. In all settings considered, sRRR has better power to detect deleterious genetic variants compared to MULM. Important issues concerning model selection and connections to existing latent variable models are also discussed. This work shows that sRRR offers a promising alternative for detecting brain-wide, genome-wide associations.

Joint prediction of multiple scores captures better individual traits from brain images

“…Reduced rank regression 120 (Vounou et al, 2010;Izenman, 1975) is a related multi-output linear-regression. Unlike CCA, it does not discard explained variance during model fitting.…”

Predicting White Matter Integrity from Multiple Common Genetic Variants

“…In addition to voxelwise maps of FA, tract-and fiber-based measures from diffusion imaging may also be considered as predictive outputs. Such measures, along with multivariate methods that simultaneously consider not only multiple genes, but also multiple voxels (Vounou et al, 2010;Hibar et al, 2011b;Wan et al, 2011) may help provide more statistical power. For instance, our voxelwise, multilocus model improved only slightly beyond the 2-SNP model with polymorphisms in NTRK1 and CLU.…”

Joint sparse canonical correlation analysis for detecting differential imaging genetics modules

“…To address this problem, penalized CCA and related methods were introduced by employing sparse penalties to select a small number of features. Examples include sparse CCA (Parkhomenko et al, 2009;Witten and Tibshirani, 2009), sparse PLS (Chun and Keles¸, 2010) and sparse reduced rank regression (Vounou et al, 2010), which have been demonstrated to be effective in detecting multivariate genomic and brain imaging associations (Grellmann et al, 2015;Liu and Calhoun, 2014). To incorporate biological prior knowledge and data structures to guide the search of associations, group SCCA (Lin et al, 2014) and network-guided sparse reduced rank regression (Wang et al, 2014) were proposed, which can further improve variable selection.…”