CCLLM is a graph-based LLM framework designed to identify conservative topological patterns of cell type combinations as CC motifs from cellular communities in SRT data.

CCLLM constructs a cellular community by cell spatial coordinates, where nodes represent individual cells and edges represent hypothetical spatial relations. The graph structure of the cellular community is then encoded a descriptive prompt template consisting of system instruction, graph structure, task description, and output format. CCLLM leverages prompt engineering to capture contextual information from the cellular community, and is fine-tuned using Low-Rank Adaptation (LoRA) to enable CC motifs identification and provide end-to-end biological interpretation.

These instructions guide you on how to run CCLLM locally.

git clone https://github.com/Springff/CCLLM.git
cd CCLLM

Create and activate a conda environment:

conda create -n myenv python=3.9
conda activate myenv

Install dependencies:

pip install -r requirements.txt

Important: You need to obtain a license for the Llama3 model and follow Meta's official instructions for downloading and configuring it.

Hugging Face: Llama3-8b-Instruct

Please consult the official Llama3 documentation for the correct download and usage instructions.

Place the downloaded Llama3 model in the llama3_models directory (recommended), or modify the model path in llama3_0.py.

python ./CCLLM.py

CCLLM/
├── CCLLM.py                # Main execution script
├── requirements.txt         # Python dependency file
├── README.md               # This document
├── Llama3_models/          # Directory for Llama3 models
├── CCLLM-Llama3/           # Directory for LoRA module
└── ...