A SAEM algorithm for fused lasso penalized NonLinear Mixed Effect Models: Application to group comparison in pharmacokinetics

Nonlinear mixed effect models are classical tools to analyze nonlinear longitudinal data in many fields such as population pharmacokinetics. Groups of observations are usually compared by introducing the group affiliations as binary covariates with a reference group that is stated among the groups. This approach is relatively limited as it allows only the comparison of the reference group to the others. The proposed method compares groups using a penalized likelihood approach. Groups are described by the same structural model but with parameters that are group specific. The likelihood is penalized with a fused lasso penalty that induces sparsity in the differences between groups for both fixed effects and variances of random effects. A penalized Stochastic Approximation EM algorithm is proposed that is coupled to Alternating Direction Method Multipliers to solve the maximization step. An extensive simulation study illustrates the performance of this algorithm when comparing more than two groups. Then the approach is applied to real data from two pharmacokinetic drug-drug interaction trials.

[1]  T. Ebner,et al.  The Metabolism and Disposition of the Oral Direct Thrombin Inhibitor, Dabigatran, in Humans , 2008, Drug Metabolism and Disposition.

[2]  Maud Delattre,et al.  A note on BIC in mixed-effects models , 2014 .

[3]  Eric Bair,et al.  Cross-validation for nonlinear mixed effects models , 2013, Journal of Pharmacokinetics and Pharmacodynamics.

[4]  É. Moulines,et al.  Convergence of a stochastic approximation version of the EM algorithm , 1999 .

[5]  Ana Arribas-Gil,et al.  LASSO-type estimators for semiparametric nonlinear mixed-effects models estimation , 2012, Stat. Comput..

[6]  H. Zou.  The Adaptive Lasso and Its Oracle Properties , 2006 .

[7]  É. Moulines,et al.  On stochastic proximal gradient algorithms , 2014 .

[8]  Hao Wang,et al.  Coordinate descent algorithm for covariance graphical lasso , 2014, Stat. Comput..

[9]  E. Ollier,et al.  A semi-mechanistic absorption model to evaluate drug-drug interaction with dabigatran: application with clarithromycin. , 2013, British journal of clinical pharmacology.

[10]  Patrick Danaher,et al.  The joint graphical lasso for inverse covariance estimation across multiple classes , 2011, Journal of the Royal Statistical Society. Series B, Statistical methodology.

[11]  Stephen P. Boyd,et al.  Distributed Optimization and Statistical Learning via the Alternating Direction Method of Multipliers , 2011, Found. Trends Mach. Learn..

[12]  France Mentré,et al.  The SAEM algorithm for group comparison tests in longitudinal data analysis based on non‐linear mixed‐effects model , 2007, Statistics in medicine.

[13]  Edouard Ollier,et al.  Joint estimation of $K$ related regression models with simple $L_1$-norm penalties , 2014, 1411.1594.

[14]  D. Balding,et al.  Multiple single nucleotide polymorphism analysis using penalized regression in nonlinear mixed-effect pharmacokinetic models , 2013, Pharmacogenetics and genomics.

[15]  R. Tibshirani,et al.  Least angle regression , 2004, math/0406456.

[16]  R. Tibshirani,et al.  Sparsity and smoothness via the fused lasso , 2005 .

[17]  E. Ollier,et al.  In vitro and in vivo evaluation of drug–drug interaction between dabigatran and proton pump inhibitors , 2015, Fundamental & clinical pharmacology.

[18]  Marc Lavielle,et al.  Maximum likelihood estimation in nonlinear mixed effects models , 2005, Comput. Stat. Data Anal..

[19]  Julien Mairal,et al.  Convex optimization with sparsity-inducing norms , 2011 .

[20]  H. Binder,et al.  A coordinate-wise optimization algorithm for the Fused Lasso , 2010, 1011.6409.

[21]  Franck Picard,et al.  On the robustness of the generalized fused lasso to prior specifications , 2014, Statistics and Computing.

[22]  Magali San Cristobal,et al.  Selection of fixed effects in high dimensional linear mixed models using a multicycle ECM algorithm , 2014, Comput. Stat. Data Anal..

[23]  V. Viallon,et al.  Regression modeling on stratified data: automatic and covariate-specific selection of the reference stratum with simple $L_1$-norm penalties , 2015 .

[24]  M. Lavielle,et al.  Model‐based analyses of bioequivalence crossover trials using the stochastic approximation expectation maximisation algorithm , 2011, Statistics in medicine.

[25]  Adeline Samson,et al.  Extension of the SAEM algorithm for nonlinear mixed models with 2 levels of random effects. , 2008, Biostatistics.

[26]  P. Bühlmann,et al.  Estimation for High‐Dimensional Linear Mixed‐Effects Models Using ℓ1‐Penalization , 2010, 1002.3784.

[27]  Gerhard Tutz,et al.  Regularization and model selection with categorical predictors and effect modifiers in generalized linear models , 2014 .

[28]  R. Tibshirani,et al.  Sparse estimation of a covariance matrix. , 2011, Biometrika.

[29]  M. Weiss,et al.  Modeling the Kinetics of Digoxin Absorption: Enhancement by P‐Glycoprotein Inhibition , 2012, Journal of clinical pharmacology.

[30]  Gerhard Tutz,et al.  Regularization and Model Selection with Categorial Effect Modifiers , 2010 .

[31]  Gersende Fort,et al.  On Perturbed Proximal Gradient Algorithms , 2014, J. Mach. Learn. Res..

[32]  M. De Iorio,et al.  Integrating dynamic mixed-effect modelling and penalized regression to explore genetic association with pharmacokinetics , 2015, Pharmacogenetics and genomics.

[33]  H. Bondell,et al.  Joint Variable Selection for Fixed and Random Effects in Linear Mixed‐Effects Models , 2010, Biometrics.

[34]  R. Tibshirani,et al.  Covariance‐regularized regression and classification for high dimensional problems , 2009, Journal of the Royal Statistical Society. Series B, Statistical methodology.