Development of a docking protocol against NS2B-NS3 pro ZIKA VIRUS

Zika is a neglected disease caused by the Zika virus (ZIKV) and so far does not have an approved treatment or vaccine. NS2B and NS3 are non-structural proteins that form a complex for viral multiplication making it a target of interest in drug development. The purpose of this study was to establish a robust protocol for conducting virtual screening to identify potential inhibitors of this protease. To select a structure for our study, analysis was conducted on all crystallographic structures accessible in the PDB. Emphasis was placed on global structure similarity and the active site residues. Additionally, the LUNA program was utilized to assess the protein-ligand interaction frequencies within the active site. The efficacy of the protocol was evaluated using various parameters, including docking power at 2 and 3 Å and RMSD values for crossdocking experiments. Furthermore, a ROC curve analysis was performed using two sets of ligands: decoys generated by DUDE-e and inactive compounds obtained from a high-throughput screening study (Abrams,2020). In the crossdocking experiments, we employed five different structures (5YOF, 5ZMS, 6KK2, 6L50, 7ZLC), each representing a distinct class of ligands (acyclic, macrocyclic, peptide, peptidomimetic, and fragments). Among these structures, 5YOF exhibited the best performance, with an average RMSD of 3.20 Å, and docking success rates of 23% of 42%, considering cutoff RMSD values of 2 Å and 3 Å, respectively. Regarding the ROC curve analysis, the AUC value for the database containing inactive compounds was 0.793 and the AUC for the decoys group was 0.557. The protocol was evaluated for the virtual screening of a library comprising 1080 compounds. Among these, the 150 top-ranked ligands were subjected to visual inspection, resulting in the selection of 11 ligands for subsequent experimental characterization.