Mini-project
Pharma & Life Sciences
Drug property predictor for solubility and toxicity
Predict solubility and toxicity of molecules from their structure, and show which parts of the molecule drive the result.
The brief
Many drug candidates fail late because they are toxic or do not dissolve. Using a public molecule dataset, predict solubility or toxicity from molecular fingerprints, compare models, and highlight the substructures that raise the risk. Good looks like: scaffold-based testing so the model faces new molecule families, honest accuracy, and examples a chemist can check.
What you will build
- Load a molecule property dataset
- Create fingerprints and descriptors with RDKit
- Split the data by scaffold
- Train and compare models
- Highlight substructures behind predictions
- Document limits for a chemist
Free, open data to build it with
Sign up free to get the data and links
It takes a minute with your email. AiTeky Pass (₹999/month or ₹9,999/year) is free for your first 30 days.