Data from: Personalized models of disorders of consciousness reveal complementary roles of connectivity and local parameters in diagnosis and prognosis

This folder gives the data necessary to reproduce the results from Personalized models of disorders of consciousness reveal complementary roles of connectivity and local parameters in diagnosis and prognosis (2025) published in Plos One by Lou Zonca, Anira Escrichs, Gustavo Patow, Dragana Manasova, Yonathan Sanz-Perl, Jitka Annen, Olivia Gosseries, Steven Laureys, Jacobo Diego Sitt and Gustavo Deco. The data is organized as follow The folder Latent_dimension_analysis provides the MSE from the Auto-Encoder training and the classification results from Fig 1 as well as an example of trained Auto-Encoder. The folder FC_matrices_all_subjects contains one Python Dictionary per subject with two items in each dict: 'FC_emp' gives the latent Empiarical Functional Connectivity in dimension 15 (the dimension used in all the study) and 'FC_emp_tau' gives the latent Empiarical Functional Connectivity where each tine-serie has been shifted of a small timestep tau (this second matrix is needed for the Generative effective connectivity (GEC) fitting procedure): this allows to reproduce the fitting results. The folders AHP_model_MBBs (resp. Hopf_model_MBB) give all the fitted model based biomarkers and fitting SSIM resuts from the rest of the paper, obtained with the two models presented.