📨 WeChat / feedback

---

What is it?

VenusFactory2 puts protein language models, biological databases, and an Agent orchestrator in one place: upload a sequence/structure to predict, drive multi-step analysis in natural language, or fine-tune your own models.

Try online: venusfactory.bio

---

Quick Start

Slow Hugging Face access: export HF_ENDPOINT=https://hf-mirror.com

git clone --recurse-submodules https://github.com/AI4Protein/VenusFactory2.git && cd VenusFactory2
# If you already cloned without submodules:
#   git submodule update --init --recursive
python scripts/setup_quickstart.py          # interactive one-click install
# python scripts/setup_quickstart.py -y     # CI / non-interactive

source .venv/bin/activate
python src/webui_v2.py --host 0.0.0.0 --port 7861
# → http://localhost:7861

The installer detects older setups (.venv / frontend / weights) and can reuse or wipe them. Press Enter for the recommended path.

After launch

1. Open the UI → Quick Tools work without an LLM key
2. Agent / Science Expert LLM: copy the endpoint config, pick a provider, then set the key

   cp model-config.yaml.example model-config.yaml
   # set `active: dmx` | `gpt` | `claude` | `gemini` | `deepseek` | …
   cp .env.example .env   # fill DMXAPI_API_KEY / OPENAI_API_KEY / ANTHROPIC_API_KEY / …
   

   The yaml is re-read on every page load and chat request — change active, model, or base_url and refresh. See model-config.yaml.example.
3. Optional check: python scripts/check_env.py

---

Documentation

Three layers — start at the top; dig only when needed:

中文：README_CN.md · docs/README.md

---

Citation

@article{tan2026venusfactory2,
  title={Self-evolving AI agents for protein discovery and directed evolution},
  author={Tan, Yang and Zhang, Lingrong and Li, Mingchen and Yu, Yuanxi and Zhong, Bozitao and Zhou, Bingxin and Dong, Nanqing and Hong, Liang},
  journal={arXiv preprint arXiv:2603.27303},
  year={2026}
}

@inproceedings{tan2025venusfactory,
  title={VenusFactory: An Integrated System for Protein Engineering with Data Retrieval and Language Model Fine-Tuning},
  author={Tan, Yang and Liu, Chen and Gao, Jingyuan and Banghao, Wu and Li, Mingchen and Wang, Ruilin and Zhang, Lingrong and Yu, Huiqun and Fan, Guisheng and Hong, Liang and others},
  booktitle={Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)},
  pages={230--241},
  year={2025}
}

---

License & acknowledgement

Released under the VenusFactory Non-Commercial License: free for academic / non-commercial use; commercial use needs written approval → [email protected]. See LICENSE / LICENSE_CN.md.

Maintained by Liang's Lab, Shanghai Jiao Tong University.
Website · YouTube · Issues