Integrated QSAR-Docking-MD Prioritization of Anthraquinone Scaffolds for Zika Virus NS5 MTase and RdRp: Target-Specific Lead Identification​

Sara Calvo Londono

Co-Presenters: Individual Presentation

College: Hennings College of Science Mathematics and Technology

Major: BS.BIO/CELL/MOLEC

Faculty Research Mentor: Kar, Supratik  

Abstract:

Zika virus (ZIKV) is a mosquito-borne pathogen that continues to pose health risks in regions suchas South America and parts of Asia. Infection can cause fever, rash, and joint pain, and in severecases, it has been linked to birth defects such as microcephaly in newborns. Currently, there is noapproved treatment or vaccine for ZIKV, highlighting the need for new antiviral strategies. TheZika virus NS5 protein is a validated antiviral target that combines two essential catalytic functionswithin a single polypeptide: an N-terminal methyltransferase (MTase) domain responsible forRNA capping and a C-terminal RNA-dependent RNA polymerase (RdRp) domain required forviral genome replication. Here, we present a dual-domain, target-specific in silico discoveryworkflow to identify anthraquinone-based inhibitors for each NS5 active site independently. Anexperimentally characterized anthraquinone dataset (n = 22) with anti-Zika activity was curated todevelop a QSAR model, establishing scaffold-level structure-activity relationships and enablingrapid prioritization of new analogs. Structure-based screening was performed via moleculardocking against MTase (PDB: 5WXB) and RdRp (PDB: 5WZ3) to rank binding poses andinteraction patterns within each catalytic pocket. Using the common anthraquinone core, additionalderivatives were retrieved from the COCONUT database and Mcule libraries and screened bydocking against both NS5 domains, followed by QSAR-based activity prediction to shortlistapproximately 15-20 candidates per target. Drug-likeness and ADMET profiling were then appliedto refine the hit lists and remove liabilities inconsistent with developability. Finally, the top fivecandidates per domain were advanced to molecular dynamics simulations to assess bindingstability and key residue–ligand interaction persistence within MTase and RdRp active sites. Thisintegrated ligand-based and structure-based pipeline delivers domain-resolved lead candidatesdistinctly optimized for MTase and RdRp, providing computationally supported starting points forsubsequent synthesis/biochemical validation and accelerating the discovery of NS5-directed Zikaantivirals.