Deep Learning-Based Ranking Method for Subgroup and Predictive Biomarkers Identification

Overview

DeepRAB is a deep learning-based framework designed for identifying subgroups and predictive biomarkers in precision medicine.

Features

• 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.

Installation

Prerequisites

Ensure you have the following installed:

• Python 3.7+
• R 4.0+