<p>Computational methods and artificial intelligence became realistic and affordable to accelerate the identification of effective treatment against SARS-CoV-2. The host-pathogen interactome analyzed was used to identify hub drug targets, which are NSP16-NSP10 and NSP15. Here we screened some natural peptides against NSP15 and NSP10-NSP16 by utilizing molecular docking. Further, we developed a machine learning model and trained neural network, SVM, AdaBoost, Random Forest and kNN algorithms. In the next step, molecular dynamics simulation and MMPBSA employed. Assessment of these models was based on R², and the neural network was the best model (R² = 0.987), which was selected to predict the pIC₅₀ of chosen peptides. Therefore, we assessed the pharmacokinetic properties of the peptides with the highest pIC50 through in-silico ADME calculations. Jaspamide passed successfully through docking, ML predictions and ADME filtration and was subjected to molecular dynamics simulation and MMPBSA, which are commonly performed to investigate the stability of protein-drug complexes; therefore, we ran them to examine the stability of complexed jaspamide with both NSP15 and NSP10-NSP16. Network pharmacology showed that jaspamide has two antiviral mechanisms: SARS-CoV-2 activates/modulates innate and adaptive immune responses through IFN-stimulated genes and SARS-CoV-2-host interactions. These computational findings suggest that jaspamide merits experimental evaluation as a potential dual inhibitor of NSP15 and NSP16, but in vitro and in vivo studies are needed to confirm in silico results.</p>