Anti-Parkinson's Therapeutic DiscoveryIhechu, Nancy Ifunanya (Avtor) Podlipnik, Črtomir (Mentor) Parkinson’s diseaseLRRK2 WD40 domainvirtual screeningmolecular dockingbinding free energyactive learningNext to Alzheimer's, Parkinson's ranks second among the most common neurodegenerative diseases, with over 25 million individuals projected to be living with this illness in 2050. Present treatments have been developed to mitigate the symptoms of Parkinson’s, but none have been able to prevent or slow down the disease. This study involves the collaboration of industry and academia, with the goal of discovering anti Parkinson’s therapeutic. Our approach uses hits from the scientific challenge CACHE (Critical Assessment of Computational Hit-Finding Experiments), with known binding free energy that was experimentally determined. Then, we aimed to identify molecules that can better bind to our target of interest, resulting in a potential therapeutic mechanism. Here, we explore the WD40 domain of LRRK2, a protein whose mutation is primarily associated with Parkinson’s disease. An initial binding pose was obtained using hits from the winning submissions of the challenge, and in a bid to identify a more probable binding pose, pocket identification and redocking was done. An ideal pose of the reference hit and protein complex's ABFE(Absolute Binding Free Energy) is calculated to achieve a more optimal binding affinity. Following this, analogs similar to the reference hits are retrieved by virtually screening the Enamine library, which contains commercially available compounds, and then template docking of these analogs to our reference pose isperformed. An iterative approach known as Active Machine Learning is then deployed in computing the RBFE for these analogs. Compounds with stronger binding affinities relative to the reference hit are selected for purchase and experimental validation to determine their potency. The result of this study will bridge the gap between experimental and computational applications in drug discovery by combining computational machine learning algorithms with experimental validation. This research lays yet another brick on developing more effective therapeutics targeting LRRK2 for Parkinson’s and other neurodegenerative diseases.20262026-05-22 10:55:02Magistrsko delo/naloga182772sl