Prostate cancer remains a significant oncological challenge, driven by molecular factors such as KLK3 (kallikrein-related peptidase 3). Text mining of 237,357 PubMed articles identified KLK3 as the most frequently cited protein (10,477 mentions in titles; 162,619 in abstracts), with strong co-mentions of AR, TMPRSS2, and ERG (χ2, *p* < 0.001). Structural modeling of KLK3 (PDB: 2ANY) using I-TASSER yielded a high-confidence 3D structure (C-score: 0.73), validated by Ramachandran analysis, with 99.5% of residues falling in favored regions. Phytochemical profiling of Curcuma longa revealed potent bioactive constituents, with leaf extracts showing the highest total phenolic (510.7 ± 0.07 µg/mL) and flavonoid (498.9 ± 0.05 µg/mL) content. LC-MS identified 23 bioactive compounds, which exhibited exceptional binding affinity. Virtual screening of FDA-approved drugs (- 11.8 to - 9.4 kcal/mol), food-derived compounds (- 10.0 to - 9.1 kcal/mol), and natural products (- 11.4 to - 8.8 kcal/mol) revealed significant differences in binding affinities. MK3207 showed the highest binding affinity (- 11.7 kcal/mol), followed by MolPort-039-338-696 (- 11.4 kcal/mol), with key interactions at PHE-110 and THR-167. In-silico docking shows that MK3207 exhibits the strongest binding affinity to KLK3 (- 11.7 kcal/mol), with accuracy validated by an RMSD of 0.195 Å. Pharmacokinetic and drug-likeness evaluation of MK3207 indicated moderate solubility (Log S: - 4.58 to - 5.02), high lipophilicity (consensus log Po/w: 3.39), favorable drug-likeness (no PAINS/Lipinski violations, bioavailability score: 0.55, synthetic accessibility: 5.21). MD simulations (100 ns) confirmed stable KLK3-ligand binding (final RMSD: 4.3 Å protein, 3.1 Å ligand; average RMSD: 3.99 Å C-α, 3.97 Å backbone, 5.41 Å sidechain). The complex exhibited moderate flexibility (RMSF peaks: 1-4 Å), 28.32% secondary structure, and persistent interactions (hydrophobic: VAL-49, PHE-110; hydrogen bonds: THR-167, SER-213). MM-GBSA analysis revealed strong binding energy (- 75.57 to - 66.36 kcal/mol) and consistent ligand efficiency. This study bridges computational drug discovery and phytochemical analysis, nominating Curcuma longa derivatives and MK3207 as promising KLK3 inhibitors for PC therapy.
Keywords: ADME properties; KLK3; MD simulation; Molecular docking; Phytochemical screening; Prostate cancer; Text mining.