RAGulate

Retrieval-Augmented Generation for Post-hoc Literature-Grounded Regulatory Validation

Introduction

RAGulate is a Retrieval-Augmented Generation (RAG) pipeline that integrates domain-specific large language models (LLM) with curated literature to identify, score, and validate inferred regulatory (transcription factor-target gene) interactions in their biological context.

For further information and example tutorials, please check our documentation:

If you have any questions or concerns, feel free to open an issue.

Requirements

RAGulate is implemented in the LlamaIndex framework. Running RAGulate on CUDA is highly recommended if available.

Before installing and running RAGulate, ensure you have the following libraries installed:

• PyTorch (version 2.0 or higher)
  Install with the exact command from the PyTorch “Get Started” page for your OS, Python version and (optionally) CUDA toolkit.
• NumPy (version 1.23 or higher)

You can install these dependencies using pip:

pip install torch numpy

Installation

Option 1 (Coming soon):
You can install RAGulate via pip for a lightweight installation:

pip install ragulate-bio

Option 2 (Coming soon):
Alternatively, if you want the latest, unreleased version, you can install it directly from the source on GitHub:

pip install git+https://github.com/YDaiLab/RAGulate.git

Import

import ragulate_bio as ragulate  # recommended alias

Note: The PyPI distribution is named ragulate-bio to avoid a name conflict with an unrelated project called ragulate. Always import ragulate_bio in Python (you may alias it to ragulate for convenience)

Option 3 (Coming soon):
For users who prefer Conda or Mamba for environment management, you can install RAGulate along with extra dependencies:

Conda:

conda install -c zandigohar RAGulate

Mamba:

mamba create -n RAGulate -c zandigohar RAGulate

FAQ

Q1: Do I need a GPU to run RAGulate?
No, a GPU is not required. However, using a CUDA-enabled GPU is strongly recommended for faster runs, especially with large queries.

Q2: How do I know if I can use a GPU with RAGulate?
There are two quick checks:

1. System check
   In your terminal, run nvidia-smi. If you see your GPU listed (model, memory, driver version), your machine has an NVIDIA GPU with the driver installed.

2. Python check
   In a Python shell, run:

   import torch
   print(torch.cuda.is_available())  # True means PyTorch can see your GPU
   print(torch.cuda.device_count())  # How many GPUs are usable

Q3: Can I use RAGulate with R-based tools?
RAGulate is written in Python and works directly with Numpy objects.

Q4: What if I also have another package called ragulate installed? RAGulate will warn you if it detects a conflicting installation. We recommend using a clean virtual environment to avoid import clashes.

Q5: How do I cite RAGulate?
See the Citation section below for the latest reference and preprint link.

Q6: How can I reproduce the paper’s results?
See our Reproducibility Guide for step-by-step instructions. Then run RAGulate.

Citation

⚠️ Preprint coming soon (bioRxiv, 2025)
This repository is under active development. Please cite as:
Zandigohar M, Rehman J, Dai Y. RAGulate: RAGulate: Retrieval-Augmented Generation for Post-hoc Literature-Grounded Regulatory Validation. 2025.

Development & Contact

RAGulate was developed and is actively maintained by Mehrdad Zandigohar as part of his PhD research at the University of Illinois Chicago (UIC), in the lab of Dr. Yang Dai.

📬 For private questions, please email: mzandi2@uic.edu

🤝 For collaboration inquiries, please contact PI: Dr. Yang Dai (yangdai@uic.edu)

Contributions, feature suggestions, and feedback are always welcome!

License

The code in RAGulate is licensed under the MIT License, which permits academic and commercial use, modification, and distribution.

Please note that any third-party dependencies bundled with RAGulate may have their own respective licenses.