Integrated in Silico Method for Virtual Screening and Interaction Profiling of Repurposed and Natural Inhibitor Candidates Against the Rdrp of the Newly Reported ANDV/Switzerland/Hu-3337/2026 Orthohantavirus

Introduction: The rapid release of complete viral genome sequences enables early computational prioritization of antiviral candidates before experimental screening becomes available. Objectives: This study evaluated the RNA-dependent RNA polymerase (RdRp; L protein) of the newly reported ANDV/Switzerland/Hu-3337/2026 orthohantavirus as an antiviral drug-design target. Methods: The L-segment-derived RdRp sequence was obtained from the public Virological.org report, modeled with AlphaFold, inspected against the NCBI Conserved Domain Database, and submitted to CB-Dock2 for binding-cavity prediction. The predicted pocket (X = 5.872, Y = 29.895, Z = -5.566) was used for DrugRep screening of 100 FDA-approved/repurposed and 100 natural TCMML compounds, which were then independently rescored with KDeep. Results: In the repurposed library, DB00511 and DB11581 gave the strongest DrugRep scores (-10.2), whereas DB00864 gave the highest KDeep pIC50 (7.1249); consensus ranking prioritized DB00511, DB00390 and DB01396. In the natural-product library, T0295 scored best with DrugRep (-9.8) and T8166 with KDeep (7.3537); consensus ranking prioritized T8166, T0295, T6473 and T3809. Interaction analysis supported these candidates through recurrent hydrogen-bond and hydrophobic contacts, including GLU129, SER157, THR170, ALA1463, TRP1464, ARG1544 and GLU1545. Conclusions: The combined DrugRep-KDeep-interaction workflow provides a rational computational shortlist for downstream redocking, molecular dynamics, ADMET profiling and experimental prioritization.