Abstract:
|
We propose a penalized generalized estimating equation framework to jointly model correlated bivariate binary and continuous outcomes involving multiple predictor variables. We use sparsity-inducing penalty functions to simultaneously estimate the regression effects and perform variable selection on the predictors, and use cross-validation to select the tuning parameters. We further propose a method for tuning parameter selection that can control a desired false discovery rate. Using simulation studies, we demonstrate that the proposed joint modelling approach performs better in terms of accuracy and variable selection than separate penalized regressions for the binary and continuous data. Finally, we demonstrate an application of this method with correlated binary and continuous phenotypes involving multiple genomic predictor variables.
|