Weighted Elastic Net Model for Mass Spectrometry Imaging Processing

D. Hong¹ ; F. Zhang¹

¹ Department of Mathematical Science, Middle Tennessee State University Murfreesboro, Tennessee, USA

Mathematical modelling of natural phenomena, Tome 5 (2010) no. 3, pp. 115-133.

Résumé

In proteomics study, Imaging Mass Spectrometry (IMS) is an emerging and very promising new technique for protein analysis from intact biological tissues. Though it has shown great potential and is very promising for rapid mapping of protein localization and the detection of sizeable differences in protein expression, challenges remain in data processing due to the difficulty of high dimensionality and the fact that the number of input variables in prediction model is significantly larger than the number of observations. To obtain a complete overview of IMS data and find trace features based on both spectral and spatial patterns, one faces a global optimization problem. In this paper, we propose a weighted elastic net (WEN) model based on IMS data processing needs of using both the spectral and spatial information for biomarker selection and classification. Properties including variable selection accuracy of the WEN model are discussed. Experimental IMS data analysis results show that such a model not only reduces the number of side features but also helps new biomarkers discovery.

Détail

DOI : 10.1051/mmnp/20105308

Bibliographie

[1] L. Breiman Better subset regression using the nonnegative garrote Technometrics 1995 373 384

[2] P. Chaurand, M.E. Sanders, R.A. Jensen, R.M. Caprioli Profiling and imaging proteins in tissue sections by MS Anal. Chem. 2004 86A 93A

[3] G. Chu, B. Narasimhan, R. Tibshirani, V.G. Tusher. SAM Version 1.12: user’s guide and technical document.[http://www-stat.stanford.edu/ tibs/SAM/]

[4] E. Candes, T. Tao The dantzig selector: statistical estimation when p is much larger than n Annals of Statistics 2007

[5] B. Efron, T. Hastie, R. Tibshirani Least angle regression Annals of Statistics 2004 407 499

[6] J. Fan, R. Li Variable selection via nonconcave penalized Likelihood and Its Oracle Properties Journal of the American Statistical Association 2001 1348 1360

[7] I. Frank, J. Friedman A statistical view of some chemometrics regression tools Technometrics 1993 109 148

[8]

[9] D.J. Graham, M.S. Wagner, D.G. Castner Information from complexity: challenges of TOF-SIMS data interpretation Applied surface science 2006 6860 6868

[10] P. Hall, J.S. Marron, A. Neeman Geometric representation of high dimension low sample size data J. R. Statist. Soc. B 2005

[11]

[12] A. E. Hoerl, R. W. Kennard Ridge regression: Biased estimation for nonorthogonal problems Technometrics 1970 55 67

[13] J. Huang, J. Horowitz, S. Ma Asymptotic properties of bridge estimators in sparse high-dimensional regression models Annals Statatistics 2008 587 613

[14] J. Huang, S. Ma, C. Zhang Adaptive Lasso for sparse high dimensional regression models Stat Sin 2008 1603 1618

[15] G.M. James, P. Radchenko, J. Lv DASSO: connections between the Dantzig selector and lasso J. R. Statist. Soc. B 2009

[16] J. Jia, B. Yu. On model selection consistency of the elastic net when p ≫ n. Tech. Report 756, Statistics, UC Berkeley, 2008.

[17] K. Knight, W. Fu Asymptotics for Lasso-type estimators Annals Statistics 2000 1356 1378

[18] S. Matoba, J.G. Kang, W.D. Patino, A. Wragg, M. Boehm, O. Gavrilova, P.J. Hurley, F. Bunz, P.M. Hwang P53 regulates mitochondrial respiration Science 2006 1650 1653

[19] S. Ma, J. Huang Penalized feature selection and classification in bioinformatics Brief in Bioinform. 2008 392 403

[20] A. Mayevsky Mitochondrial function and energy metabolism in cancer cells: Past overview and future perspectives Mitochondrion 2009 165 179

[21] G. Mccombie, D. Staab, M. Stoeckli, R. Knochenmuss Spatial and Spectral correlation in MALDI mass spectrometry images by clustering and multivariate analysis Anal. Chem. 2005 6118 6124

[22] N. Meinshausen, B. Yu Lasso-type recovery of sparse representations for high-dimensional data Annals of Statistics 2009 246 270

[23] H. Meistermann, J.L. Norris, H.R. Aerni, D.S. Cornett, A. Friedlein, A.R. Erskine, A. Augustin, M.C. De Vera Mudry, S. Ruepp, L. Suter, H. Langen, R.M. Caprioli, A. Ducret Biomarker discovery by imaging mass spectrometry: transthyretin is a biomarker for gentamicin-induced nephrotoxicity in rat Mol Cell Proteomics 2006 1876 1886

[24] E.R. Muir, I.J. Ndiour, N.A. Le Goasduff, R.A. Moffitt, Y. Liu, M.C. Sullards, A.H. Merrill, Y. Chen, M.D. Wang. Multivariate analysis of imaging mass spectrometry data. BIBE 2007 proceedings of the 7th IEEE international conference 472-479.

[25] R. Tibshirani Regression shrinkage and selection via the lasso J. R. Statist. Soc., Series B. 1996 267 288

[26] M. Yuan, Y. Lin On the nonnegative garrote estimator J. R. Statist. Soc. B. 2007 143 161

[27] F. Zhang, D. Hong, S. Frappier, D.S. Cornett, R.M. Caprioli. Elastic Net Based Framework for Imaging Mass Spectrometry Data Biomarker Selection and Classification. Manuscript, 2009.

[28] H. Zhang, J. Ahn, X. Lin, C. Park Gene selection using support vector machines with non-convex penalty Bioinformatics 2006 88 95

[29] P. Zhao, B. Yu On model selection consistency of lasso The Journal of Machine Learning Research 2006 2541 2563

[30] S. Zhou, S. Geer, P. Buhlmann. Adaptive lasso for high dimensional regression and gaussian graphical modeling. manuscript, 2009.

[31] H. Zou The adaptive lasso and its oracle properties Journal of the American Statistical Association 2006 1418 1429

[32] H. Zou, T. Hastie Regularization and variable selection via the elastic net J. R. Statist. Soc., B. 2005 301 320

[33] H. Zou, H. Zhang On the adaptive elastic-net with a diverging number of parameters Annals of statistics 2009 1733 1751