DeCoDE

A deep learning framework for contrastive dimensional embedding

DeCoDE is designed to identify robust, transdiagnostic brain–behavior dimensions and neurophysiological biotypes by integrating functional connectivity with behavioral measures.

---

Yong Jiao, Xiaoyu Tong, Gregory A. Fonzo, Ian H. Gotlib, Kilian M. Pohl, Theodore D. Satterthwaite, Jing Jiang, Yu Zhang Deep Learning of Brain–Behavior Dimensions Identifies Transdiagnostic Biotypes in Youth with ADHD and Anxiety Disorders

Project Structure

.
├── assets/
│   └── flowchart.png                   # Method overview figure
├── models.py                           # cVAE and GCCA model components
├── run_DeCoDE_pipeline.py              # Executable workflow
├── requirements.txt                    # Python dependencies
└── README.md                           # This file

Dataset

• Adolescent Brain Cognitive Development (ABCD) Study: https://abcdstudy.org/
• Healthy Brain Network (HBN): https://fcon_1000.projects.nitrc.org/indi/cmi_healthy_brain_network

Dependencies

This project is implemented in Python 3.14.2. Install dependencies via:

pip install -r requirements.txt

Quick Start

Run the DeCoDE workflow on random matrices:

python run_DeCoDE_pipeline.py \
  --epochs 5 \
  --batch-size 256 \
  --hidden-dims 256 \
  --latent-dim 32 \
  --alpha 15 \
  --gamma 5 \
  --device gpu