DeepRAB is a deep learning-based framework designed for identifying subgroups and predictive biomarkers in precision medicine.
- DeepRAB Framework: Implements the core DeepRAB model for identifying subgroups and predictive biomarkers.
- Causal Forest Framework: Integrates the Causal Forest (CF) model for estimating conditional average treatment effects (CATE) as a comparison.
- XGBoost with Modified Loss Function: A customized version of XGBoost tailored for biomarker identification, incorporating an A-learning loss function.
- Linear Regression Models: Implements linear regression with both modified outcomes and modified covariates.
Ensure you have the following installed:
- Python 3.7+
- R 4.0+