skills.rest
shushuzn
shushuznCOMMUNITY
Explore all skills by this author →
GitHub

esm

Generate and analyze protein sequences, structures, and annotations with ESM3.

4|Updated Mar 2, 2026
Search Tags:#embeddings#structure prediction#protein design#esm3#function annotation#forge api#inverse folding
One-click install
npx skills add https://github.com/shushuzn/Rairos --skill esm-shushuzn
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: esm
Source: https://github.com/shushuzn/Rairos/tree/main/skills/esm
Command: npx skills add https://github.com/shushuzn/Rairos --skill esm-shushuzn

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Help scientists generate, complete, and optimize protein sequences and predict protein structure or function without manually engineering features from scratch.

Core Features & Use Cases

  • ESM3 protein generation across modalities: Generate protein sequence, structure (coordinates), and function annotations using iterative masked modeling, including inverse folding from target structure inputs.
  • ESM C embeddings for representation learning: Compute efficient protein embeddings for similarity, clustering, classification, and downstream predictive modeling.
  • Forge API for scalable inference: Use cloud-based ESM3 models for larger and production-grade generation workloads when local hardware is insufficient.
  • Use case: You need a pipeline that designs multiple protein variants, predicts their structures, then clusters the results using embeddings to select diverse candidates for further experimental validation.

Quick Start

Use the esm skill to generate a completed protein sequence from a partially masked sequence like MPRT___KEND.

Frequently Asked Questions about esm

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

FAQPage Schema
How do I design and complete protein sequences from partial structures?▼

To design and complete protein sequences, you input a partially masked sequence using underscore tokens. ESM3 then uses iterative masked modeling to generate the missing sequence, structure, and function annotations.

What is inverse folding and how does it work for protein design?▼

Inverse folding for protein design involves generating a protein sequence from a target 3D structure input. ESM3 accomplishes this by processing the structure coordinates and producing the corresponding amino acid sequence.

Can I compute protein embeddings for clustering and similarity search?▼

Yes, you can compute protein embeddings for clustering and similarity search. ESM C generates efficient representations that directly support classification, similarity search, and downstream predictive modeling workflows.

How do I run large-scale protein generation when local hardware is insufficient?▼

You can run large-scale protein generation using the Forge API. It provides cloud-based ESM3 execution for production-grade workloads, requiring YAML configuration and optional API authentication for cloud scalability.

Do I need YAML configuration to predict protein structure and function?▼

Yes, YAML configuration is required to predict protein structure and function. You must configure model selection, format protein inputs with masked positions, and set generation parameters by modality track with step count and sampling temperature.

Related Skills

Explore More Education & Research Skills→
silversteinsilverstein
21

esm

Generate protein sequences, predict structures, and compute embeddings with ESM3 and ESM C.

Community
Advanced
Estrella-231Estrella-231

esm

Generate protein sequences, predict structures, and extract embeddings with ESM3 and ESM C.

Community
Advanced
yf8578yf8578
1

esm

Generate protein sequences, predict structures, and extract embeddings using ESM3 and ESM C.

Community
Advanced
robotlearning123robotlearning123

esm

Generate protein sequences, structures, and functions using ESM3 and ESM C models.

Community
Advanced