Low-Bias Evaluation of Drug-Target Interaction

This project is in collaboration with AstraZeneca.

Drug-target interaction (DTI) prediction is important in drug discovery and chemogenomics studies. Machine learning, particularly deep learning, has advanced this area significantly over the past few years. However, a significant gap between the performance reported in academic papers and that in practical drug discovery settings, e.g. the random-split-based evaluation strategy tends to be too optimistic in estimating the prediction performance in real-world settings. Such performance gap is largely due to hidden data bias in experimental datasets and inappropriate data split.

In this project, we construct a low-bias DTI dataset and study more challenging data split strategies to improve performance evaluation for real-world settings. Specifically, we study the data bias in a popular DTI dataset, BindingDB, and re-evaluate the prediction performance of three state-of-the-art deep learning models using five different data split strategies: random split, cold drug split, scaffold split, and two hierarchical-clustering-based splits. In addition, we comprehensively examine six performance metrics. Our experimental results confirm the overoptimism of the popular random split and show that hierarchical-clustering-based splits are far more challenging and can provide potentially more useful assessment of model generalizability in real-world DTI prediction settings.

Peizhen Bai
Peizhen Bai
PhD Student
Bino John
Bino John
Global Team Leader at AstraZeneca
Haiping Lu
Haiping Lu
Professor of Machine Learning, Head of AI Research Engineering, and Turing Academic Lead

I am a Professor of Machine Learning. I develop translational AI technologies for better analysing multimodal data in healthcare and beyond.

Related