batch_cell_typing

Analyze the effects of intensity batch correction algorithms on cell typing models (MAPS & XGBoost) for 3 different methods:

1. COMBAT
2. MICHEALSON
3. Z-score normalization

Getting Started

1. Git pull the MAPS repo + Install related dependencies
2. Replace the trainer.py script in MAPS/maps/cell_phenotyping/trainer.py with the script in this repo
   • This script only changes the input for the model.fit function to take in dataframes instead of a path
   • Implemented just for ease of use and batch running
3. Run maps_main.ipynb

TO DO

[] Clean up the ipynb script