BBBper: A Machine Learning-based Online Tool for Blood-brain Barrier (BBB) Permeability Prediction
Pawan Kumar, Vandana Saini, Dinesh Gupta, Pooja A. Chawla, Pooja A. Chawla, Ajit Kumar
Abstract
INTRODUCTION: BBB limits the permeability of external compounds by 98% to maintain and regulate brain homeostasis. Hence, BBB permeability prediction is vital to predict the activity of a drug-like substance. AIM: Neuronal disorders have affected more than 15% of the world's population, signifying the importance of continued design and development of drugs that can cross the Blood-Brain Barrier (BBB). OBJECTIVE: Here, we report about developing BBBper (Blood-Brain Barrier permeability prediction) using machine learning tool. METHODS: A supervised machine learning-based online tool, based on physicochemical parameters to predict the BBB permeability of given chemical compounds was developed. The user-end webpage was developed in HTML and linked with back-end server by a python script to run user queries and results. RESULTS: BBBper uses a random forest algorithm at the back end, showing 97% accuracy on the external dataset, compared to 70-92% accuracy of currently available web-based BBB permeability prediction tools. CONCLUSION: The BBBper web tool is freely available at http://bbbper.mdu.ac.in.