Benchmarking Accuracy and Generalizability of Four Graph Neural Networks Using Large In Vitro ADME Datasets from Different Chemical Spaces

Computer Science

• Traditional Model 100%
• Deep Learning Model 100%
• Chemical Space 100%
• Graph Neural Network 100%
• Generalizability 100%
• Graph Attention Network 100%
• Message Passing 33%
• Neural Network Model 33%
• Machine Learning Approach 33%
• Network Message 33%
• Task Graph 33%
• Graph Convolutional Network 33%
• Neural Network 33%

Pharmacology, Toxicology and Pharmaceutical Science

• Absorption Distribution Metabolism Excretion 100%

Keyphrases

• Single-task Model 33%
• Message Passing Neural Network 33%