Patients with Parkinson’s Disease (PD) may have very different patterns of progression, corresponding to distinct disease subtypes. Here, we describe quantitatively the overall pattern of progression in subgroups of PD by using a Bayesian non-linear mixed effect model that describes the continuous progression of biomarkers at both population and individual level. This approach allows to model variability in progression patterns and disease stage between patients. We analyzed two subgroups of patients, with (RBD+) and without sleep disorders (RBD-), that are known to present different patterns of progression [1]. We compared the two groups by extracting the ordering of abnormalities that occurred over the disease course, and by studying their disease onset and speed of progression.