Skip to content

BioEmu

Generate protein structure samples with deep learning

Visit official site

What BioEmu is

BioEmu is described as a Microsoft deep learning model that generates protein structure samples to simulate equilibrium ensembles, aimed at biochemistry, structural biology and drug design.

What you can do with it

  • Generate structure samples for a given protein sequence
  • Run scalable equilibrium ensemble simulations
  • Combine samples with side-chain reconstruction and MD relaxation

Application scenarios

  • Researchers study protein folding with generated samples
  • Drug designers look for possible binding sites in the ensemble
  • Educators explain dynamic protein structures in class

Getting started

  • Clone the repository and run the install script for a Conda environment
  • Use the sampling script with a sequence and a sample count
  • Add side-chain and MD relaxation tools if needed
  • Post-process the structures with the matching module

Pros & cons

✓ What we like

  • Complex biomolecular systems are within its stated scope
  • Side-chain and MD relaxation tools can be added

! What to watch out for

  • The described output is structure samples of an ensemble
  • Setup uses a Conda environment built by an install script

FAQ

What does it generate?

The material states protein structure samples.

Where are the weights hosted?

The material says model parameters can be downloaded from Hugging Face.

Last reviewed: 2026-11-15

More LLM API platform tools

View all →

How we review