Predicting AXL Tyrosine Kinase Inhibitor Potency Using Machine Learning with Interpretable Insights for Cancer Drug Discovery DOI
Teuku Rizky Noviandy,

Ghifari Maulana Idroes,

Essy Harnelly

et al.

Heca Journal of Applied Sciences, Journal Year: 2025, Volume and Issue: 3(1), P. 17 - 29

Published: March 15, 2025

AXL tyrosine kinase plays a critical role in cancer progression, metastasis, and therapy resistance, making it promising target for therapeutic intervention. However, traditional drug discovery methods developing inhibitors are resource-intensive, time-consuming, often fail to provide detailed insights into molecular determinants of potency. To address this gap, we applied machine learning techniques, including Random Forest, Gradient Boosting, Support Vector Regression, Decision Tree models, predict the potency (pIC50) using dataset 972 compounds with 550 descriptors. Our results demonstrate that Forest model outperformed others an R² 0.703, MAE 0.553, RMSE 0.720, PCC 0.841, showcasing strong predictive accuracy. SHAP analysis identified features, such as RNCG TopoPSA(NO), key contributors inhibitor potency, providing interpretable structure-activity relationships. These findings highlight potential accelerate identification optimization inhibitors, bridging gap between computational predictions rational design paving way effective therapeutics.

Language: Английский

Predicting AXL Tyrosine Kinase Inhibitor Potency Using Machine Learning with Interpretable Insights for Cancer Drug Discovery DOI
Teuku Rizky Noviandy,

Ghifari Maulana Idroes,

Essy Harnelly

et al.

Heca Journal of Applied Sciences, Journal Year: 2025, Volume and Issue: 3(1), P. 17 - 29

Published: March 15, 2025

AXL tyrosine kinase plays a critical role in cancer progression, metastasis, and therapy resistance, making it promising target for therapeutic intervention. However, traditional drug discovery methods developing inhibitors are resource-intensive, time-consuming, often fail to provide detailed insights into molecular determinants of potency. To address this gap, we applied machine learning techniques, including Random Forest, Gradient Boosting, Support Vector Regression, Decision Tree models, predict the potency (pIC50) using dataset 972 compounds with 550 descriptors. Our results demonstrate that Forest model outperformed others an R² 0.703, MAE 0.553, RMSE 0.720, PCC 0.841, showcasing strong predictive accuracy. SHAP analysis identified features, such as RNCG TopoPSA(NO), key contributors inhibitor potency, providing interpretable structure-activity relationships. These findings highlight potential accelerate identification optimization inhibitors, bridging gap between computational predictions rational design paving way effective therapeutics.

Language: Английский

Citations

0