Targeting mitogen-activated protein kinase 4 for liver regeneration through QSAR-based virtual screening and unbiased MD simulation.

Mitogen-Activated Protein Kinase Kinase 4 (MKK4) regulates key signaling pathways including JNK and p38 MAPK, influencing apoptosis, proliferation, and stress responses. Dysregulation of MKK4 is implicated in aggressive cancers and impaired liver regeneration, making it a compelling therapeutic target. This study employed an integrative computational strategy to identify potent MKK4 inhibitors. A curated dataset of 2936 compounds was analyzed using twelve fingerprint descriptors to develop robust QSAR models with Random Forest. The CDK extended fingerprint model showed excellent performance (R2 = 0.96 ± 0.00; Q2_CV = 0.78 ± 0.03; Q2_Ext = 0.80 ± 0.02) and was used to virtually screen the ZINC, NPA, and FDA libraries. Top hits FDA3675381442 and NPC100204 were prioritized based on their predicted inhibitory activity and favorable drug-like properties. Molecular docking revealed strong binding affinities and key interactions with critical MKK4 active site residues, including Ser312, Ala112, Gly109, Tyr113, and Lys231. Both compounds exhibited superior docking scores (-9.16 kcal/mol for FDA3675381442 and - 8.71 kcal/mol for NPC100204) compared to the reference ligand AMP-PNP. Molecular dynamics simulations over 250 ns confirmed the structural stability of the complexes, as indicated by low RMSD/RMSF values and compact radius of gyration. MM/PBSA binding free energy calculations further supported their high affinity (-43.90 ± 0.11 kcal/mol and - 45.70 ± 0.14 kcal/mol, respectively). Collectively, these results suggest that FDA3675381442 and NPC100204 are promising MKK4 inhibitors with strong and stable interactions, warranting further investigation as therapeutic candidates for cancer and liver regeneration.