scllm

A Python package for annotating single-cell RNA sequencing data using Large Language Models.

Overview: LangChain 🤝 scanpy

scllm leverages the power of Large Language Models to automatically annotate cell types in single-cell RNA sequencing data. It integrates seamlessly with scanpy and provides an intuitive interface for cell type annotation, factor analalysis and differential expression analysis based on marker gene expression.

Installation

You can install scllm using pip:

pip install scllm

Or using uv 🚀:

uv pip install scllm

Quick Start

import os

# Enter your API key for ChatGPT
os.environ["OPENAI_API_KEY"] = "Enter your API key here."

import scanpy as sc
import scllm

# Load your data
adata = sc.read_h5ad('your_data.h5ad')

# Perform clustering if not already done
sc.tl.leiden(adata)

# Initialize your LLM (example with OpenAI)
from langchain_openai import ChatOpenAI
llm = ChatOpenAI(temperature=0.0, model="gpt-4o-mini")

# Annotate clusters
adata = scllm.annotate_cluster(adata, 'leiden', llm)

# Access annotations
print(adata.obs['leiden_annotated'])

Features

• Automatic cell type annotation using LLMs
• Seamless integration with scanpy
• Support for multiple LLM providers
• Interactive Jupyter notebook examples
• Customizable annotation parameters

Documentation

For detailed documentation and examples, visit our documentation page.

Check out our example notebooks:

• Cell Type Annotation

Requirements

• Python ≥ 3.10
• scanpy ≥ 1.11.0
• langchain ≥ 0.3.7
• And other dependencies listed in pyproject.toml

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Citation

If you use scllm in your research, please cite:

@software{vohringer2024scllm,
  author = {Vöhringer, Harald},
  title = {scllm: Single-Cell Annotation with Large Language Models},
  year = {2024},
  publisher = {GitHub},
  url = {https://github.com/sagar87/scllm}
}

Contact

Harald Vöhringer - harald.voeh@gmail.com

Project Link: https://github.com/sagar87/scllm