<![CDATA[Melanoma is an aggressive skin cancer with high mortality rates, necessitating the development of novel therapeutic agents. This study aims to identify potential anti-melanoma candidates from triterpenoid derivatives through an integrated chemoinformatics approach. A dataset of 35 triterpenoid compounds was analyzed using KNIME to build a Quantitative Structure-Activity Relationship (QSAR) model. The model demonstrated good predictive ability with R² = 0.842, RMSE = 0.38, and MAE = 0.30, validated through 5-fold cross-validation (Q² = 0.78). QSAR analysis identified five best candidates: cucurbitacin B (IC₅₀ = 0.015 µM), betulin (15.61 µM), lupeol (66.59 µM), oleanolic acid (75 µM), and ursolic acid (75 µM). Lipinski's Rule of Five analysis revealed LogP violations in most compounds, though these are common for natural triterpenoids and can be addressed through formulation strategies. Molecular docking showed oleanolic acid had the best binding affinity against BCL-2 (-8.2 kcal/mol), while lupeol showed the highest affinity against BRAF (-8.9 kcal/mol). Based on the balance of predicted activity, pharmacokinetic profiles, and binding affinity, oleanolic acid and lupeol emerged as the most promising candidates for further experimental validation. This integrative approach demonstrates the efficiency of computational screening in prioritizing lead compounds for melanoma drug discovery.]]>
Copyrights © 2026
