This project explores the use of machine learning techniques to identify potential drug repurposing opportunities. The objective is to leverage existing drug–target, drug–gene, and drug–enzyme relationships to discover new therapeutic applications for approved drugs.
- Data integration from multiple biological and chemical sources
- Feature engineering on drug–drug and drug–target relationships
- Supervised machine learning models for candidate prediction
- Comparative analysis of model performance
- Random Forest
- Support Vector Machines (SVM)
- Ensemble learning techniques
- Machine learning models identified promising drug candidates for further investigation
- Feature-based approaches demonstrated effectiveness in capturing biological relationships
Python, Scikit-learn, Pandas, NumPy
This project was developed as part of applied machine learning research and demonstrates the use of ML techniques beyond traditional security and energy domains.
#Data sources
drugbank.ca
http://www.orphadata.org/cgi-bin/index.php
https://rarediseases.info.nih.gov/about-gard/pages/23/about-gard