bioemu

Sample protein sequences to generate backbone conformation ensembles with YAML-configurable denoiser pipelines.

2|Updated May 12, 2026
One-click install
npx skills add https://github.com/LiorZ/protein-design-skills --skill bioemu
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: bioemu
Source: https://github.com/LiorZ/protein-design-skills/tree/main/skills/bioemu
Command: npx skills add https://github.com/LiorZ/protein-design-skills --skill bioemu

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

BioEmu enables sampling an ensemble of backbone conformations for a protein monomer from its sequence, delivering an ensemble instead of a single predicted structure, at a fraction of MD cost.

Core Features & Use Cases

  • Ensemble sampling of monomer backbones with optional steering (SMC/FKC) to improve realism or target a collective variable; side-chain reconstruction via HPacker; optional short MD relaxation; and file outputs that support downstream analysis.
  • Use cases include exploring cryptic pockets, estimating folding/free-energy landscapes (ΔG_fold), and performing conformational analyses to guide design and validation workflows.
  • Compatible with Bring-Your-Own-MSA or standard ColabFold embedding pipelines, with resumable sampling and steerable configurations for reproducibility.

Quick Start

Run a quick test by sampling a small ensemble from a sequence using a modest number of samples and inspect the topology.pdb and samples.xtc outputs.

Frequently Asked Questions about bioemu

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a protein ensemble for conformational sampling instead of a single predicted structure?

Protein ensemble sampling produces an array of backbone conformations rather than a single static structure by sampling a monomer's sequence. This diffusion-based approach captures conformational variability for downstream analysis.

What is the best way to estimate folding free-energy landscapes without running expensive molecular dynamics simulations?

Estimating folding free-energy landscapes requires generating a protein backbone ensemble at a fraction of MD cost. This diffusion-based conformational sampling provides the structural variance needed to calculate ΔG_fold.

Can I use a diffusion model to identify cryptic pockets in monomeric proteins?

Yes, you can identify cryptic pockets by sampling an ensemble of backbone conformations for a monomeric protein. Exploring this structural variability reveals hidden binding sites that static predictions often miss.

Do I need Linux and Python to run diffusion-based conformational sampling for protein ensembles?

Yes, Linux and Python are required to run this diffusion-based conformational sampling pipeline. Optional GPU acceleration is supported to improve processing speed during the generation of protein ensembles.

How do I steer protein backbone sampling toward a specific collective variable?

You can steer protein backbone sampling toward a collective variable using configurable SMC or FKC methods within the denoiser pipeline. This improves realism or targets specific conformational states during ensemble generation.

What output files are generated for downstream conformational analysis of protein ensembles?

The sampling process outputs samples.xtc and topology.pdb files, alongside batch caches. These outputs support trajectory visualization and downstream conformational analysis with resumable and reproducible semantics.