Related Experiment Video
Updated: May 5, 2026

A Bilingual Computational Workflow for Identifying Potential PLK1 Inhibitors in American Sign Language and English
Published on: April 3, 2026
Computational Investigation of Novel pUL56 Ligands Using Docking and Molecular Dynamics with Preliminary Cytotoxicity
Viktoria Feoktistova1, Samson Olusegun Afolabi1, Artem M Klabukov2
1Infochemistry Scientific Center, Information Technologies, Mechanics and Optics (ITMO) University, Saint Petersburg 191002, Russia.
Abstract:
Human cytomegalovirus (HCMV) remains a significant cause of morbidity in immunocompromised patients, necessitating the development of improved antivirals. Using an integrated in silico and in vitro approach, we identified a novel ligand (NL) as a letermovir analog with enhanced binding affinity and reduced cytotoxicity. A pUL56 terminase subunit model generated with AlphaFold 3 was used for the virtual screening of a 15,000-compound library. Among the 73 candidates with structural similarity to letermovir (Tanimoto ≥ 0.6), NL exhibited superior predicted binding affinity (ΔGbind = -10.7 kcal/mol). In silico toxicity prediction (ProTox 3.0) classified NL as having low toxicity (class 4, LD50 ≈ 1000 mg/kg), which was confirmed in vitro, where NL demonstrated 158-fold less toxic (CC50 = 2.69 mg/mL) in MRC-5 cells than letermovir (0.017 mg/mL). Molecular dynamics simulations over 500 ns revealed that the pUL56-NL complex forms a more thermodynamically stable interaction, with a lower calculated free energy of binding (MMGBSA: -40.89 ± 7.40 kcal/mol vs. -32.76 ± 4.96 kcal/mol) and a narrower free energy landscape. These results establish NL as a promising, low-cytotoxicity candidate with enhanced target engagement, warranting further investigation as a potential anti-HCMV therapeutic.

