Parkinson’s disease (PD) is a degressive, neurodegenerative disease that affects approximately four million people worldwide. The disease is characterized by the progressive loss of midbrain dopaminergic (DAergic) neurons, which are highly related with the motor control. As the disease progresses, movement disorders appear such as tremor, rigidity, and bradykinesia, but also disorders in speech and neuropsychiatric disturbances occur.Current therapies for PD focus on symptomatic treatment, while pharmacological methods to prevent or delay the degeneration of neurons have not been discovered yet. The Nurr1 nuclear receptor, which is expressed predominantly in the substantia nigra of the midbrain, has emerged as a target for the treatment of Parkinson’s disease due to its neuroprotective action and contribution in DAergic neuron development. It has been shown that partial loss of Nurr1 function in people due to mutations leads to neuronal death. Thus, the reinforcement of Nurr1 operation via the discovery of novel potent agonists is imperative. Unfortunately, the accomplishment of this task is complicated as Nurr1 ligand binding domain (LBD) lacks a cavity for ligand binding due to the tight packing chains from several hydrophobic amino-acids in the region normally occupied by ligands in other nuclear receptors. However, the activation of Nurr1 can be feasible through heterodimer formation with Retinoid X Receptors (RXR) and especially with RXRα, which are all capable of binding ligands and therefore, mediate Nurr1 expression in midbrain. Therefore, we seek here to identify potent binders of RXRα as a means to increase Nurr1 levels. Based on the fact that multiple RXRα receptor conformations exist depending on binding of RXRα to different heterodimerization partners, we aim to increase the specificity of identified binders for the heterodimer Nurr1/RXRα. For this purpose, we describe here a new computational protocol for the selection of RXRα receptor structures that is used to perform Structure-Based Virtual Screening (SBVS) calculations for the discovery of NURR1 activators. In our study, we developed a computational protocol, where the choice of RXRα conformations for performing the SBVS is based on four criteria: (a) Pairwise comparison of the receptor conformations according to RMSD calculations, (b) analysis and clustering of RXRα structures comparing the binding-site shape and volume using SiteMap, (c) docking of a small-database of known actives for a specific heterodimer partner to the resulting shape-diverse subset of binding sites from (a) and (b) using Glide 5.8 SP and XP, and (d) retrieving representative protein conformations for the structure of interest from MD simulations using GROMACS. Virtual Screening was performed on three different subsets of RXRα receptor conformations, based on their binding to different heterodimerization partners. The final RXRα receptors to be used in SBVS were selected as mentioned above aiming to enhance the success rate and the selectivity of the hits. The Maybridge Hitfinder and Zinc databases were used in this SBVS exercise by first applying the SP filter on the full database and then the XP filter on the top 10,000 compounds of the Maybridge database and the top 40,000 compounds of the ZINC database. Compounds were selected as follows: Molecules that scored high when docked in the RXRα protein ensemble that bind to the heterodimer partner of interest and at the same time scored low for RXRα structures that bind to heterodimer partners of no interest, were selected in order to achieve selectivity. The efficiect selection was also based on their different orientation at the binding site of the various RXRα structures and different interactions with specific surrounding residues in order to maximize their selectivity potential. Finally, a post-processing step was imposed to the top-scoring compounds by using Chembioserver and FAF-Drugs2 filtering tools as well as pharmacological property prediction with the QikProp software. In vitro agonism of these compounds is still pending experimental testing. The workflow of this protocol is shown in Fig. 1. Figure 1: SBVS protocol developed for the discovery of novel selective Nurr1/RXRα agonists. / --