MixRx uses a Large Language Model (LLM) to analyze drug interactions and recommend an Additive, Synergistic, or Antagonistic drug, given a multi-drug patient history.

1. Run generate_validation_data.py

• drug_combo_dict.json contains all the existing pairs of drug synergy calculations that exist
• drug_data contains the raw data pulled from synergxdb
• valid_drug_combinations.csv contains random possible drug combinations where every pair of drugs have synergy data

2. Run generate_model_data.py (for the model's prompts)

• output_prompts_contexts turns the input data for the model into a string based prompt

3. Run synergy.ipynb

• validate the model's output via the automated synergy calculator