Accurate ADMET Prediction with XGBoost

Hao Tian, Rajas Ketkar, Peng Tao

2022-04-15Molecular Property Prediction Drug Discovery Prediction TDC ADMET Benchmarking Group

Abstract

The absorption, distribution, metabolism, excretion, and toxicity (ADMET) properties are important in drug discovery as they define efficacy and safety. In this work, we applied an ensemble of features, including fingerprints and descriptors, and a tree-based machine learning model, extreme gradient boosting, for accurate ADMET prediction. Our model performs well in the Therapeutics Data Commons ADMET benchmark group. For 22 tasks, our model is ranked first in 18 tasks and top 3 in 21 tasks. The trained machine learning models are integrated in ADMETboost, a web server that is publicly available at https://ai-druglab.smu.edu/admet.

Results

Task	Dataset	Metric	Value	Model
Drug Discovery	tdcommons	TDC.AMES	0.859	XGBoost
Drug Discovery	tdcommons	TDC.BBB_Martins	0.905	XGBoost
Drug Discovery	tdcommons	TDC.Bioavailability_Ma	0.7	XGBoost
Drug Discovery	tdcommons	TDC.CYP2C9_Inhibition_Veith	0.877	XGBoost
Drug Discovery	tdcommons	TDC.CYP2C9_Substrate_CarbonMangels	0.68	XGBoost
Drug Discovery	tdcommons	TDC.CYP2D6_Inhibition_Veith	0.794	XGBoost
Drug Discovery	tdcommons	TDC.CYP2D6_Substrate_CarbonMangels	0.387	XGBoost
Drug Discovery	tdcommons	TDC.CYP3A4_Inhibition_Veith	0.721	XGBoost
Drug Discovery	tdcommons	TDC.CYP3A4_Substrate_CarbonMangels	0.648	XGBoost
Drug Discovery	tdcommons	TDC.Caco2_Wang	0.288	XGBoost
Drug Discovery	tdcommons	TDC.Clearance_Hepatocyte_AZ	0.587	XGBoost
Drug Discovery	tdcommons	TDC.Clearance_Microsome_AZ	0.42	XGBoost
Drug Discovery	tdcommons	TDC.DILI	0.933	XGBoost
Drug Discovery	tdcommons	TDC.HIA_Hou	0.987	XGBoost
Drug Discovery	tdcommons	TDC.Half_Life_Obach	0.396	XGBoost
Drug Discovery	tdcommons	TDC.LD50_Zhu	0.602	XGBoost
Drug Discovery	tdcommons	TDC.Lipophilicity_AstraZeneca	0.533	XGBoost
Drug Discovery	tdcommons	TDC.PPBR_AZ	8.251	XGBoost
Drug Discovery	tdcommons	TDC.Pgp_Broccatelli	0.911	XGBoost
Drug Discovery	tdcommons	TDC.Solubility_AqSolDB	0.727	XGBoost
Drug Discovery	tdcommons	TDC.VDss_Lombardo	0.612	XGBoost
Drug Discovery	tdcommons	TDC.hERG	0.806	XGBoost
Molecular Property Prediction	BBBP	ROC-AUC	90.5	XGBoost
Atomistic Description	BBBP	ROC-AUC	90.5	XGBoost
Therapeutics Data Commons	tdcommons	TDC.AMES	0.859	XGBoost
Therapeutics Data Commons	tdcommons	TDC.BBB_Martins	0.905	XGBoost
Therapeutics Data Commons	tdcommons	TDC.Bioavailability_Ma	0.7	XGBoost
Therapeutics Data Commons	tdcommons	TDC.CYP2C9_Inhibition_Veith	0.877	XGBoost
Therapeutics Data Commons	tdcommons	TDC.CYP2C9_Substrate_CarbonMangels	0.68	XGBoost
Therapeutics Data Commons	tdcommons	TDC.CYP2D6_Inhibition_Veith	0.794	XGBoost
Therapeutics Data Commons	tdcommons	TDC.CYP2D6_Substrate_CarbonMangels	0.387	XGBoost
Therapeutics Data Commons	tdcommons	TDC.CYP3A4_Inhibition_Veith	0.721	XGBoost
Therapeutics Data Commons	tdcommons	TDC.CYP3A4_Substrate_CarbonMangels	0.648	XGBoost
Therapeutics Data Commons	tdcommons	TDC.Caco2_Wang	0.288	XGBoost
Therapeutics Data Commons	tdcommons	TDC.Clearance_Hepatocyte_AZ	0.587	XGBoost
Therapeutics Data Commons	tdcommons	TDC.Clearance_Microsome_AZ	0.42	XGBoost
Therapeutics Data Commons	tdcommons	TDC.DILI	0.933	XGBoost
Therapeutics Data Commons	tdcommons	TDC.HIA_Hou	0.987	XGBoost
Therapeutics Data Commons	tdcommons	TDC.Half_Life_Obach	0.396	XGBoost
Therapeutics Data Commons	tdcommons	TDC.LD50_Zhu	0.602	XGBoost
Therapeutics Data Commons	tdcommons	TDC.Lipophilicity_AstraZeneca	0.533	XGBoost
Therapeutics Data Commons	tdcommons	TDC.PPBR_AZ	8.251	XGBoost
Therapeutics Data Commons	tdcommons	TDC.Pgp_Broccatelli	0.911	XGBoost
Therapeutics Data Commons	tdcommons	TDC.Solubility_AqSolDB	0.727	XGBoost
Therapeutics Data Commons	tdcommons	TDC.VDss_Lombardo	0.612	XGBoost
Therapeutics Data Commons	tdcommons	TDC.hERG	0.806	XGBoost

Accurate ADMET Prediction with XGBoost

Abstract

Results

Related Papers

Accurate ADMET Prediction with XGBoost

Abstract

Results

Related Papers