SP 2 - CKDNapp

Mathematical modeling for CKDNapp

This subproject will develop the mathematical models that enter CKDNapp. We will (1) design risk scores to predict adverse events in patients with chronic kidney disease (CKD), (2) establish classifiers for CKD staging refinement, (3) generate metabolic models and facilitate/verify their use by routine diagnostic assays, (4) provide comprehensive mathematical modeling of CKD by mixed graphical models (MGMs) taking into account the complex interplay between all patient parameters, and (5) enrich CKDNapp models with time-course information. We calculate these CKDNapp models based on different biomedical and demographic variables, as well as untargeteted metabolomics data of the German Chronic Kidney Disease (GCKD) study employing state-of-the-art machine learning methods.

Network illustration of a mixed graphical model, which uncovers statistical associations between different variables measured within the GCKD study. Circles or rectangles represent individual demographic, drug treatment, clinical chemistry, and metabolic variables, edges between two variables represent an association between these respective variables.