BBBP Prediction Project

This project aims to predict Blood-Brain Barrier Penetration (BBBP) using machine learning models based on molecular structures and features.

Overview

The BBBP dataset consists of molecular features extracted from chemical compounds along with their BBBP values. The objective is to build predictive models that determine whether a compound can penetrate the blood-brain barrier.

Columns:

• Smiles: SMILES representation of chemical compounds
• Name: Name of chemical compounds
• p_np: BBBP classification (1: Penetrates, 0: Doesn't Penetrate)

Setup

1. Clone the repository.

git clone https://github.com/jaywyawhare/Molecular-Classification

2. Install dependencies.

pip install -r requirements.txt

Model Architecture

Future Improvements

• Hyperparameter tuning for models to enhance performance.
• Incorporate additional molecular descriptors for better predictive power.
• Explore deep learning architectures for improved accuracy.

License

This project is licensed under DBaJ-NC-CFL. Refer to LICENSE for more details.