Identifying quantitative trait loci via group-sparse multitask regression and feature selection: an imaging genetics study of the ADNI cohort

Wang, Hua; Nie, Feiping; Huang, Heng; Kim, Sungeun; Nho, Kwangsik; Risacher, Shannon L.; Saykin, Andrew J.; Shen, Li

doi:10.1093/bioinformatics/btr649

Cited by 158 publications

(178 citation statements)

References 24 publications

Supporting

Mentioning

174

Contrasting

Unclassified

Order By: Relevance

“…Thus, the predictors within each group are more homogeneous and should be considered together for predicting the response variables. To solve this issue, we have proposed a novel Group-Sparse Multi-task Regression and Feature Selection (GSMuRFS) method [19] to exploit the interrelated structures within and between the predictor and response variables (Fig. 1).…”

Section: G-smurfsmentioning

confidence: 99%

“…To address this issue, we propose to employ a new structured sparse learning model called G-SMuRFS (Group-Sparse Multi-task Regression and Feature Selection) [19] for multivariate regression of cognitive scores on neuroimaging data. Motivated by Lasso [16] and group Lasso [24], G-SMuRFS was proposed by us in an imaging genetic application [19], where an 2,1 -norm was used to bundle all the outcome imaging measures together and a group 2,1 -norm (G 2,1 -norm) was used to model the block structure within genetic predictors.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Multimodal Neuroimaging Predictors for Cognitive Performance Using Structured Sparse Learning

Yan

Risacher

Kim

et al. 2012

Multimodal Brain Image Analysis

Self Cite

View full text Add to dashboard Cite

Regression models have been widely studied to investigate whether multimodal neuroimaging measures can be used as effective biomarkers for predicting cognitive outcomes in the study of Alzheimer's Disease (AD). Most existing models overlook the interrelated structures either within neuroimaging measures or between cognitive outcomes, and thus may have limited power to yield optimal solutions. To address this issue, we propose to incorporate an L21 norm and/or a group L21 norm (G21 norm) in the regression models. Using ADNI-1 and ADNI-GO/2 data, we apply these models to examining the ability of structural MRI and AV-45 PET scans for predicting cognitive measures including ADAS and RAVLT scores. We focus our analyses on the participants with mild cognitive impairment (MCI), a prodromal stage of AD, in order to identify useful patterns for early detection. Compared with traditional linear and ridge regression methods, these new models not only demonstrate superior and more stable predictive performances, but also identify a small set of imaging markers that are biologically meaningful.

show abstract

Section: G-smurfsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Multimodal Neuroimaging Predictors for Cognitive Performance Using Structured Sparse Learning

Yan

Risacher

Kim

et al. 2012

Multimodal Brain Image Analysis

Self Cite

View full text Add to dashboard Cite

show abstract

“…Wang et al, 2011;Greenlaw et al, 2017). However, these methods become computationally intractable when analyzing data with tens of thousands of genotyped variants.…”

Section: Introductionmentioning

confidence: 99%

Multivariate association between single-nucleotide polymorphisms in Alzgene linkage regions and structural changes in the brain: discovery, refinement and validation

Szefer

Nathoo

et al. 2017

Statistical Applications in Genetics and Molecular Biology

View full text Add to dashboard Cite

Using publicly-available data from the Alzheimer's Disease Neuroimaging Initiative, we investigate the joint association between single-nucleotide polymorphisms (SNPs) in previously established linkage regions for Alzheimer's disease (AD) and rates of decline in brain structure. In an initial, discovery stage of analysis, we applied a weighted RV test to assess the association between 75,845 SNPs in the Alzgene linkage regions and rates of change in structural MRI measurements for 56 brain regions affected by AD, in 632 subjects. After confirming association, we selected refined lists of 1694 and 22 SNPs via a bootstrapenhanced sparse canonical correlation analysis. In a final, validation stage, we confirmed association between the refined list of 1694 SNPs and the imaging phenotypes in an independent data set. Genes corresponding to priority SNPs having the highest contribution in the validation data have previously been implicated or hypothesized to be implicated in AD, including GCLC, IDE, and STAMBP1andFAS. Though the effect sizes of the 1694 SNPs in the priority set are likely small, further investigation within this set may advance understanding of the missing heritability in AD. Our analysis addresses challenges in current imaging-genetics studies such as biased sampling designs and high-dimensional data with low association signal.

show abstract

“…For example, the functionality of the human brain typically involves more than one cerebral component. By investigating each individual regional brain phenotype separately will lead to a loss of informative interactions between them [25]. Furthermore, in the presence of highly correlated features Lasso tends to only select one of these features, resulting in suboptimal performance [33].…”

Section: Introductionmentioning

confidence: 99%

Discriminative Lasso

et al. 2016

View full text Add to dashboard Cite

Lasso-type variable selection has been demonstrated to be effective in handling high-dimensional data. From the biological perspective, traditional Lassotype models are capable of learning which stimuli are valuable while ignoring the many that are not, and thus perform feature selection. Traditional Lasso has the tendency to over-emphasize sparsity and to overlook the correlations between features. These drawbacks have been demonstrated to be critical in limiting its performance on real-world feature selection problems. Although some work has considered the problem of correlation, the issue of discriminative ability resulting from sparsity has been overlooked. To overcome this shortcoming, we propose a discriminative Lasso (referred to as dLasso) in which sparsity and correlation are jointly considered. Specifically, the new method can select features (or stimuli) that are correlated more strongly with the response but are less correlated with each other. Moreover, an efficient alternating direction method of multipliers (ADMM) is presented to solve the resulting sparse non-convex optimization problem. Extensive experiments on different datasets show that although our proposed model is not a convex problem, it outperforms both its approximately convex counterparts and a number of state-of-the-art feature selection methods.

show abstract

Identifying quantitative trait loci via group-sparse multitask regression and feature selection: an imaging genetics study of the ADNI cohort

Abstract: Software is publicly available at: http://ranger.uta.edu/%7eheng/imaging-genetics/.

Cited by 158 publications

References 24 publications

Multimodal Neuroimaging Predictors for Cognitive Performance Using Structured Sparse Learning

Multimodal Neuroimaging Predictors for Cognitive Performance Using Structured Sparse Learning

Multivariate association between single-nucleotide polymorphisms in Alzgene linkage regions and structural changes in the brain: discovery, refinement and validation

Discriminative Lasso

Contact Info

Product

Resources

About