Parkinson’s disease (PD) is the second most common neurodegenerative disorder. The current evidence shows that there is still a marked heterogeneity in the subtyping of PD using both clinical and data-driven approaches. The most commonly, patients can be divided into groups based on motor symptoms such as tremor dominant (TD), akinesia-rigidity (AR), and postural instability and gait disorder (PIGD). Each phenotype demonstrates different disease progression, symptom severity, and clinical features. Further response to deep brain stimulation (DBS) therapy may also be phenotype dependent. Using peri-operative electrophysiology (e.g., LFP, SUA, EEG), neuroimaging (e.g., fMRI), signal processing, machine-learning decoding, and neuromodulation (DBS), we are working to characterize the neural signatures of PD phenotypes and develop more adaptive therapies.