B1242
Title: Jewel 2.0: An improved joint estimation method for multiple Gaussian graphical models
Authors: Anna Plaksienko - University of Oslo (Norway) [presenting]
Claudia Angelini - Istituto per le Applicazioni del Calcolo CNR-Napoli (Italy)
Daniela De Canditiis - CNR (Italy)
Abstract: An upgraded method, Jewel 2.0, is presented for the joint estimation of Gaussian graphical models from multiple sources. The first version allowed the estimation of graphical models (graphs of conditional dependencies between variables) given several datasets (coming from various conditions) under the assumption that all the connections are the same across conditions. The second version has two penalties in its regression-based minimization problem, thus modelling commonality and class-specific differences in graph structures. Moreover, Jewel 2.0 better estimates graphs with hubs, making this new approach more appealing for biological data applications. A novel stability selection procedure is presented in the multiple graphs setting to reduce the number of false positives in the estimated graphs. The method is implemented in the new version of the R package Jewel.
