VITRO_VIVO_Correlation

ChEMBL-PubChem-ProteinAtlas Integration Tool

an attempt to generate examples of invitro vs invivo data correlation for drugs.

Setup Instructions

pip install -r requirements.txt

2. Prepare Data Files

Make sure you have unzipped the protein_atlas.tsv file in the same directory as your Streamlit app.

3. Run the Application

streamlit run streamlined_app.py

Features

1. Compound Search

• Input Types Supported:
  • SMILES strings (e.g., CCO for ethanol)
  • PubChem CIDs (e.g., 2244 for aspirin)
  • ChEMBL IDs (e.g., CHEMBL25 for aspirin)

API Rate Limits

The app makes calls to:

• ChEMBL REST API: Generally no strict limits, but be respectful
• PubChem REST API: Limit to ~5 requests per second

Data Sources

1. ChEMBL Database: Bioactivity data for drug discovery
2. PubChem Database: Chemical information and biological activities
3. Human Protein Atlas: Protein expression and localization data

Example Compounds to Test

• Aspirin: CID=2244, ChEMBL=CHEMBL25, SMILES=CC(=O)OC1=CC=CC=C1C(=O)O
• Caffeine: CID=2519, ChEMBL=CHEMBL113, SMILES=CN1C=NC2=C1C(=O)N(C(=O)N2C)C
• Ibuprofen: CID=3672, ChEMBL=CHEMBL521, SMILES=CC(C)CC1=CC=C(C=C1)C(C)C(=O)O