🧪 What Drug Is This Most Like?

A small tool that takes a SMILES string (a text representation of a molecule) and tells you which well-known drugs it structurally resembles most, using Tanimoto similarity on Morgan fingerprints (RDKit).

Built as a fun summer side project to play around with cheminformatics before starting an MSc - not a rigorous screening tool, just a way to explore how structural similarity search works.

Example

Paste a SMILES string: CC(=O)Nc1ccc(O)cc1

🔬 Your molecule - MW: 151.2, LogP: 1.35

Top 5 closest known drugs:

1. Paracetamol     100.0%  ████████████████████
   ↳ The one everyone has in their cupboard.
2. Aspirin          22.2%  ████
   ↳ The OG painkiller, been around since 1897.
3. Amoxicillin      21.8%  ████
   ↳ Common penicillin-type antibiotic.
4. Warfarin         18.8%  ███
   ↳ Blood thinner, also rat poison lol.
5. Ibuprofen        18.4%  ███
   ↳ NSAID, your gym-injury best friend.

How it works

1. Converts each molecule's SMILES string into a Morgan fingerprint (a bit-vector representation of its structural features).
2. Compares the input molecule's fingerprint against a small library of ~20 well-known drugs using Tanimoto similarity.
3. Ranks and returns the closest matches.

Setup

pip install rdkit

Usage

Command line:

python what_drug_is_this.py "CC(=O)Nc1ccc(O)cc1"

Or run it without an argument and paste a SMILES string when prompted:

python what_drug_is_this.py

Web app (Streamlit):

pip install streamlit
streamlit run app.py

Try it with these SMILES strings

Drug	SMILES
Aspirin	CC(=O)Oc1ccccc1C(=O)O
Caffeine	Cn1cnc2c1c(=O)n(C)c(=O)n2C
Paracetamol	CC(=O)Nc1ccc(O)cc1

Or find SMILES for any other drug on PubChem

• search a drug name and copy the "Canonical SMILES" field.

Notes

• Similarity scores are based purely on 2D structural fingerprints - they don't account for 3D shape, binding affinity, or actual pharmacological effect. A high score means "structurally similar," not "does the same thing."
• The drug reference library is small and hand-picked for fun/recognisability, not comprehensive.

Built with

• RDKit - cheminformatics toolkit
• Streamlit - web app framework (optional UI layer)