Biostatistics, Data Integration and AI

Basic support in statistical data analysis is provided. We welcome any request dealing with the implementation of standard (e.g., hypothesis testing, ANOVA, mixed-models, clustering) or advanced (e.g., non-linear regression and classification, feature and representation learning) statistical methods, to assess, validate, correlate or explore biomedical data (typically, clinical examinations, cell imaging, electrophysiology, genomics or metabolomics data acquisitions).

In addition, we design and apply specific methods for the integrative analysis of multimodal and high-dimensional data (e.g., genetics/multiomics data, neuroimaging data, clinical observations). In particular, a versatile framework called Regularized Generalized Canonical Correlation Analysis (RGCCA), and its sparse counterpart SGCCA, are dedicated to the analysis of data sets structured in blocks of variables. In partnership with CentraleSupelec, we develop new components for this package (e.g., for better handling missing values) and implement graphical interfaces to operate the methods and visualize the results.