Biocatalysis Assistant

Biocatalysis Assistant is a Python-based chatbot framework designed to automate and streamline bioinformatics and biocatalysis tasks. By integrating advanced language models and a dynamic set of tools, it simplifies complex processes in computational biology and enzyme engineering, making research more efficient and accessible.

---

Table of Contents

1. Features
2. Tools
3. Installation
   • Docker Setup and Usage
   • Local Installation
4. Usage
   • Command Line Interface (CLI)
   • Streamlit Web Application
5. References
6. License
7. Support

---

Features

• 🗣️ Interactive Chatbot Interface: Engage with the assistant via a user-friendly command-line or web interface.
• 🔧 Dynamic Tool Integration: Seamlessly run bioinformatics and biocatalysis tools.
• 🤖 Language Model Integration: Leverage state-of-the-art language models for analysis and decision-making.
• 🔗 Extensible Architecture: Easily add new tools and functionalities.
• 📈 Optimization Capabilities: Optimize enzyme sequences for enhanced catalytic activity.
• 🔍 Comprehensive Analysis: Perform detailed bioinformatics analyses, including binding site extraction and reaction element analysis.

---

Tools

The Biocatalysis Assistant integrates the following tools for bioinformatics and biocatalysis tasks:

• ExtractBindingSites: Extract binding sites from enzyme sequences.
• GetElementsOfReaction: Analyze and extract elements from biochemical reactions.
• Blastp: Perform BLASTP searches for protein sequence analysis.
• OptimizeEnzymeSequences: Optimize enzyme sequences for improved catalytic activity.
• FindPDBStructure: Search for related protein structures in the PDB database.
• DownloadPDBStructure: Download PDB structure files.
• Mutagenesis: Perform in silico mutagenesis on protein sequences.
• MDSimulation: Run molecular dynamics simulations.

Note:

For a complete description of the tool and its usage, refer to tools_description.md.

---

Installation

Docker Setup and Usage

The Biocatalysis Assistant can be run using Docker, which simplifies the setup process by encapsulating all dependencies in a container. If you use Docker, you can skip the local installation steps below.

1. Build the Docker Image

docker build -t lmabc .

2. Run the Docker Container

docker run -it -p 8501:8501 --env-file .env lmabc lmabc-app

• -it: Runs the container interactively.
• -p 8501:8501: Maps port 8501 on your host to the container.
• --env-file .env: Passes environment variables.

3. Mount Local Directories

To persist data or use local files:

docker run -it -p 8501:8501 --env-file .env --volume ${HOME}/my-cache:/app/.lmabc lmabc lmabc-app

4. Access the Streamlit Web App

Open your browser and navigate to:

http://localhost:8501

5. Use the CLI Inside the Container

docker run -it --env-file .env lmabc:latest lmabc --help

6. Example Commands

• Run the Streamlit Web App:

  docker run -it -p 8501:8501 --env-file .env lmabc:latest lmabc-app

• Use the CLI:

  docker run -it --env-file .env lmabc:latest lmabc

• Run with Local Cache:

  docker run -it -p 8501:8501 --env-file .env --volume ${HOME}/my-cache:/app/.lmabc lmabc:latest lmabc-app

---

Local Installation

If you prefer to run the Biocatalysis Assistant locally, follow these steps.

Prerequisites

• Python: Version 3.10 or higher. Download Python.
• GROMACS: A molecular dynamics simulation package. Install GROMACS.

---

1. Clone the Repository

Clone the repository to your local machine:

git clone [email protected]:GT4SD/lm-assistant-for-biocatalysis.git
cd lm-assistant-for-biocatalysis

---

2. Set Up the Environment File

1. Create a .env file in the project root directory.
2. Add your credentials (e.g., for Hugging Face API):

   HUGGINGFACEHUB_API_TOKEN=your_key_here

3. Important: Ensure .env is added to .gitignore to protect sensitive information.

---

3. Install Dependencies

Install dependencies using uv:

pip install uv
uv pip install -e .

---

4. Set Up Tools and Databases

Run the setup script to configure molecular dynamics files, RXNAAMapper, enzyme optimization models, and BLAST databases. Use the --blastdb option to specify which BLAST database(s) to download:

• Default: Downloads Swissprot database.
• Custom: Pass a specific database name (e.g., nr).
• All: Pass all to download a predefined set of databases.

For macOS Users with Homebrew

bash tools_setup/setup.sh --blastdb <db_name|all>

For Other Operating Systems or Without Homebrew

Skip Minio Client installation:

bash tools_setup/setup.sh --skip-mc --blastdb <db_name|all>

---

5. Install Additional Tools

• PyMOL: Install via Conda.
• GROMACS: Follow the official installation guide.

---

Usage

Command Line Interface (CLI)

To start the Biocatalysis Assistant using the CLI:

1. Ensure you have completed the installation process using Uv.
2. Open your terminal.
3. Run the following command:

lmabc --help

---

Streamlit Web Application

To run the Biocatalysis Assistant using the Streamlit web interface:

1. Ensure you have completed the installation process using Uv.
2. Open your terminal.
3. Run the following command:

lmabc-app

---

References

If you use lmabc in your projects, please cite:

@software{LMABC,
  author = {Yves Gaetan Nana Teukam, Francesca Grisoni, Matteo Manica},
  month = {10},
  title = {The biocatalysis assistant: a language model agent for biocatalysis (lmabc)},
  url = {https://github.com/GT4SD/lm-assistant-for-biocatalysis},
  version = {main},
  year = {2024}
}

---

License

The lmabc codebase is under the MIT license. For individual model usage, refer to the licenses of the original packages.

---

Support

For issues or questions, please open an issue in the GitHub repository.

---