Regularized Multivariate Regression for Identifying Master Predictors with Application to Integrative Genomics Study of Breast Cancer

By Jie Peng, Ji Zhu, A. Bergamaschi, Wonshik Han, Dong-Young Noh, Jonathan R. Pollack, Pei Wang. Annals of Applied Statistics, 2008

305 citations recorded by the source corpus.

In this paper, we propose a new method remMap — REgularized Multivariate regression for identifying MAster Predictors — for fitting multivariate response regression models under the high-dimension-low-sample-size setting. remMap is motivated by investigating the regulatory relationships among different biological molecules based on multiple types of high dimensional genomic data. Particularly, we are interested in studying the influence of DNA copy number alterations on RNA transcript levels. For this purpose, we model the dependence of the RNA expression levels on DNA copy numbers through multivariate linear regressions and utilize proper regularization to deal with the high dimensionality as well as to incorporate desired network structures. Criteria for selecting the tuning parameters are also discussed. The performance of the proposed method is illustrated through extensive simulation studies. Finally, remMap is applied to a breast cancer study, in which genome wide RNA transcript levels and DNA copy numbers were measured for 172 tumor samples. We identify a trans-hub region in cytoband 17q12–q21, whose amplification influences the RNA expression levels of more than 30 unlinked genes. These findings may lead to a better understanding of breast cancer pathology.

Subfields: Molecular Biology

View Paper: https://www.semanticscholar.org/paper/23be7d2d5472797576f0d0840ac4c46b0996125f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

Citations in the corpus

SLEP: Sparse Learning with Efficient Projections (2011, methodology): “Sparse Group LassoAs an extension of Lasso and group Lasso, the sparse group Las so penalty [14, 51] yields a solution that achieves the within- and between- gro up sparsity simultaneously.”
Collaborative Sparse Regression for Hyperspectral Unmixing (2014, methodology): “Combined 2 ,1 and 1,1 norm regularizers (i.e., sums of 2 or 1 norms, respectively) are used in [34] to deal with multivariate response regression problems in which the data are high-dimensional, but a low number of samples is available.”
A Selective Review of Group Selection in High-Dimensional Models. (2012, methodology): “Some examples of using group selection methods in these applications include Wei and Li (2007), Jacob, Obozinski and Vert (2009), Ma and Huang (2009), Ma, Huang and Moran (2009), Ma, Huang and Song (2010), Ma et al. (2011), Pan, Xie and Shen (2010) and Peng et al. (2010), among others.”
A note on the group lasso and a sparse group lasso (2010, background): “A similar penalty involving both group lasso and lasso terms is discussed in Peng et al. (2009).”
Feature Selection (2016, methodology): “Once Group Lasso selects a group, all the features in the selected group will be kept.”
Robust Joint Graph Sparse Coding for Unsupervised Spectral Feature Selection (2017, unknown): “For example, an 1-norm regularizer [43], [44], [48] leads to the element sparsity and an 2,1-norm regularizer [49] results in the group sparsity, while a mixed-norm regularizer combining an 1-norm with an 2,1-norm [50] produces the mixed sparsity.”
Identification of Partially Linear Structure in Additive Models with an Application to Gene Expression Prediction from Sequences (2012, methodology): “After the introduction of lasso in Tibshirani (1996), the penalization approach for variable selection has become increasingly popular (Zou and Hastie, 2005; Zhao and Yu, 2006; Zhang and Huang, 2008; Bickel, Ritov, and Tsybakov, 2009; Wu and Liu, 2009; Yuan, Joseph, and Zou, 2009; Peng et al., 2010)”
Block-Row Sparse Multiview Multilabel Learning for Image Classification (2016, methodology): “It is clear that the proposed algorithm was sensitive to the parameters’ setting, which was the same as the conclusion in [11] and [56].”
Reduced rank stochastic regression with a sparse singular value decomposition (2012, background): “Recently, as high dimensional data become increasingly common, another approach of dimension reduction, through utilizing sparsity-inducing regularization techniques under a multivariate regression framework, has emerged (Turlach et al., 2005; Obozinski et al., 2008; Peng et al., 2010).”
Sparse Multivariate Regression With Covariance Estimation (2010, methodology): “Imposing sparsity in ˆ B for the purposes of identifying “master pre-dictors” was proposed by Peng et al. (2009), who applied a combined constraint function where C 1 ( B ) = j,k | b jk | , the lasso constraint (Tibshirani, 1996) on the entries of B and P The ﬁrst constraint introduces sparsity in…”