Regularized quantile regression under heterogeneous sparsity with application to quantitative genetic traits

Abstract

Genetic studies often involve quantitative traits. Identifying genetic features that influence quantitative traits can help to uncover the etiology of diseases. Quantile regression method considers the conditional quantiles of the response variable, and is able to characterize the underlying regression structure in a more comprehensive manner. On the other hand, genetic studies often involve high-dimensional genomic features, and the underlying regression structure may be heterogeneous in terms of both effect sizes and sparsity. To account for the potential genetic heterogeneity, including the heterogeneous sparsity, a regularized quantile regression method is introduced. The theoretical property of the proposed method is investigated, and its performance is examined through a series of simulation studies. A real dataset is analyzed to demonstrate the application of the proposed method.

Analysis of CO2 emissions and energy consumption by sources in MENA countries: evidence from quantile regressions

“…The model is also advantageous in the sense of no conditional assumption about the distribution of the data (Sherwood and Wang 2016). Lastly, the model minimizes the sum of squares of the residuals by adopting the linear programming method (He et al 2016). Despite the listed advantages, the model is ineffective in terms of estimating a large number of fixed effects and exhibits incidental parameter problems when T is small.…”

Stable Anatomy Detection in Multimodal Imaging Through Sparse Group Regularization: A Comparative Study of Iron Accumulation in the Aging Brain

“…Implications of these challenges are apparent in the drawbacks (Davatzikos, 2004 ) of voxel-based analyses (Ashburner and Friston, 2000 ) and motivate many of the dimension reduction and regularization techniques commonly applied to neuroimaging data. These methods include pattern identification and discrimination (Fan et al, 2007 ), which rely on sample sizes not feasible in small neuroimaging studies, and sparse regularization for feature extraction or selection (Batmanghelich et al, 2011 ; Sabuncu and Van Leemput, 2012 ; He et al, 2016 ; Yu et al, 2019 ; Su et al, 2020 ). Despite successful applications of the latter to medical data (Krishnapuram et al, 2005 ; Zou and Hastie, 2005 ; Ryali et al, 2010 ), these approaches are unstable with respect to selected or extracted features.…”

“…We formulate this problem in terms of sparse regression with L 1 (lasso), TV, and L 2 group lasso penalties, with the last of these combining information across multiple modalities. These penalties have been widely used in the development of robust methods for medical data (He et al, 2016 ; Yu et al, 2019 ). Together, the penalties regularize spatial effect estimates for better interpretability, both within and between modalities.…”

Sparse Multicategory Generalized Distance Weighted Discrimination in Ultra-High Dimensions

“…To overcome this weakness in this paper, we developed a multicategory generalized DWD method that is capable of performing variable selection and classification simultaneously. Our approach incorporates sparsity and group structure information via the sparse group lasso penalty (see [ 20 , 21 , 22 , 23 , 24 ]).…”