Skip to content

ProteinForge

Protein design tools like RFdiffusion are incredibly powerful but quite tricky to use for non researchers.

Code Demo Cerebral Valley

Core trick

A Gemini agent converts plain-English protein requests into RFdiffusion jobs, then renders PyMOL images to critique and iterate.

Tags

scienceagents

Description

Protein design tools like RFdiffusion are incredibly powerful but quite tricky to use for non researchers. We built a chat interface where you describe a protein in plain English — "make something that looks like a donut" — and a Gemini agent translates that into RFdiffusion runs, iterates on the designs using its multimodal vision capabilities (renders .pngs with PyMol), and shows you the results in a 3D viewer with a timeline slider. The backend is a FastAPI server that manages jobs and runs Gemini as a subagent. The frontend is a Next.js chat app with gemini from openrouter and Vercel AI SDK streaming.
michaelwaves/proteinforge★ 00 forksPythonNOASSERTIONCreated Mar 21, 2026 · last push Mar 22, 2026

README excerpt

# RF*diffusion*

<!--
<img width="1115" alt="Screen Shot 2023-01-19 at 5 56 33 PM" src="https://user-images.githubusercontent.com/56419265/213588200-f8f44dba-276e-4dd2-b844-15acc441458d.png">
-->
<p align="center">
  <img src="./img/diffusion_protein_gradient_2.jpg" alt="alt text" width="1100px" align="middle"/>
</p>

*Image: Ian C. Haydon / UW Institute for Protein Design*

## Description

RFdiffusion is an open source method for structure generation, with or without conditional information (a motif, target etc). It can perform a whole range of protein design challenges as we have outlined in [the RFdiffusion paper](https://www.biorxiv.org/content/10.1101/2022.12.09.519842v1).

**Things Diffusion can do**
- Motif Scaffolding
- Unconditional protein generation
- Symmetric unconditional generation (cyclic, dihedral and tetrahedral symmetries currently implemented, more coming!)
- Symmetric motif scaffolding
- Binder design
- Design diversification ("partial diffusion", sampling around a design)

## [Documentation](https://sites.google.com/omsf.io/rfdiffusion/overview)
View the RFdiffusion documentation resource maintained by Rosetta Commons [here](https://sites.google.com/omsf.io/rfdiffusion/overview). 

----

# Table of contents

- [RF*diffusion*](#rfdiffusion)
  - [Description](#description)
- [Table of contents](#table-of-contents)
- [Getting started / installation](#getting-started--installation)
    - [Conda Install SE3-Transformer](#conda-install-se3-transformer)
    - [

Team