Camilo Broc1,2, Therese Truong3,4, Benoit Liquet5,6. 1. LIST, CEA, Laboratory for Data Sciences and Decision (Digiteo), Gif-sur-Yvette, France. camilo.broc@gmail.com. 2. CNRS, Laboratoire de Mathématiques et de leurs Applications de PAU E2S UPPA, Pau, France. camilo.broc@gmail.com. 3. UVSQ, Inserm, CESP, Université Paris-Saclay, 94807, Villejuif, France. 4. Institut Gustave Roussy, 94805, Villejuif, France. 5. CNRS, Laboratoire de Mathématiques et de leurs Applications de PAU E2S UPPA, Pau, France. 6. Department of Mathematics and Statistics, Macquarie University, Sydney, Australia.
Abstract
BACKGROUND: The increasing number of genome-wide association studies (GWAS) has revealed several loci that are associated to multiple distinct phenotypes, suggesting the existence of pleiotropic effects. Highlighting these cross-phenotype genetic associations could help to identify and understand common biological mechanisms underlying some diseases. Common approaches test the association between genetic variants and multiple traits at the SNP level. In this paper, we propose a novel gene- and a pathway-level approach in the case where several independent GWAS on independent traits are available. The method is based on a generalization of the sparse group Partial Least Squares (sgPLS) to take into account groups of variables, and a Lasso penalization that links all independent data sets. This method, called joint-sgPLS, is able to convincingly detect signal at the variable level and at the group level. RESULTS: Our method has the advantage to propose a global readable model while coping with the architecture of data. It can outperform traditional methods and provides a wider insight in terms of a priori information. We compared the performance of the proposed method to other benchmark methods on simulated data and gave an example of application on real data with the aim to highlight common susceptibility variants to breast and thyroid cancers. CONCLUSION: The joint-sgPLS shows interesting properties for detecting a signal. As an extension of the PLS, the method is suited for data with a large number of variables. The choice of Lasso penalization copes with architectures of groups of variables and observations sets. Furthermore, although the method has been applied to a genetic study, its formulation is adapted to any data with high number of variables and an exposed a priori architecture in other application fields.
BACKGROUND: The increasing number of genome-wide association studies (GWAS) has revealed several loci that are associated to multiple distinct phenotypes, suggesting the existence of pleiotropic effects. Highlighting these cross-phenotype genetic associations could help to identify and understand common biological mechanisms underlying some diseases. Common approaches test the association between genetic variants and multiple traits at the SNP level. In this paper, we propose a novel gene- and a pathway-level approach in the case where several independent GWAS on independent traits are available. The method is based on a generalization of the sparse group Partial Least Squares (sgPLS) to take into account groups of variables, and a Lasso penalization that links all independent data sets. This method, called joint-sgPLS, is able to convincingly detect signal at the variable level and at the group level. RESULTS: Our method has the advantage to propose a global readable model while coping with the architecture of data. It can outperform traditional methods and provides a wider insight in terms of a priori information. We compared the performance of the proposed method to other benchmark methods on simulated data and gave an example of application on real data with the aim to highlight common susceptibility variants to breast and thyroid cancers. CONCLUSION: The joint-sgPLS shows interesting properties for detecting a signal. As an extension of the PLS, the method is suited for data with a large number of variables. The choice of Lasso penalization copes with architectures of groups of variables and observations sets. Furthermore, although the method has been applied to a genetic study, its formulation is adapted to any data with high number of variables and an exposed a priori architecture in other application fields.
Entities:
Keywords:
Genetic epidemiology; High dimensional data; Lasso Penalization; Meta-analysis; Oncology; Partial Least Square; Pathway analysis; Pleiotropy; Sparse methods; Variable selection
Authors: Samsiddhi Bhattacharjee; Preetha Rajaraman; Kevin B Jacobs; William A Wheeler; Beatrice S Melin; Patricia Hartge; Meredith Yeager; Charles C Chung; Stephen J Chanock; Nilanjan Chatterjee Journal: Am J Hum Genet Date: 2012-05-04 Impact factor: 11.025
Authors: Babak Shahbaba; Robert Tibshirani; Catherine M Shachaf; Sylvia K Plevritis Journal: J R Stat Soc Ser C Appl Stat Date: 2011-08-01 Impact factor: 1.864
Authors: Benjamin M Neale; Manuel A Rivas; Benjamin F Voight; David Altshuler; Bernie Devlin; Marju Orho-Melander; Sekar Kathiresan; Shaun M Purcell; Kathryn Roeder; Mark J Daly Journal: PLoS Genet Date: 2011-03-03 Impact factor: 5.917