What is the difference between ESM3 and ESM C?

ESM3 is a multimodal generative model designed for sequence, structure, and function prediction and design, while ESM C is optimized for creating efficient, high-quality protein embeddings and representations.

How does ESM handle structure prediction?

ESM3 uses its structure track to predict 3D coordinates from a sequence or perform inverse folding to design a sequence for a given PDB structure.

Can I use this skill for local protein modeling?

Yes, the skill supports local usage of open-weight models like ESM3-sm-open-v1 and ESMC-300m using PyTorch and GPU acceleration.

Is an API key required?

While local models can be used for free, accessing larger models like ESM3-Large or ESM3-Medium requires a Forge API token from EvolutionaryScale.

ESM Protein Engineering Toolkit

Name: ESM Protein Engineering Toolkit
Author: Microck

byMicrock

•

Ciencia de Datos y ML

Empowers protein research and design using state-of-the-art ESM3 and ESM C language models for sequence generation, structure prediction, and representation learning.

Acerca de

The ESM skill integrates advanced Evolutionary Scale Modeling tools into Claude, enabling researchers and developers to design novel proteins, predict 3D structures, and generate high-quality embeddings. It supports the multimodal ESM3 model for generative tasks across sequence, structure, and function tracks, as well as the efficient ESM C family for representation learning. Whether performing inverse folding, function-conditioned generation, or large-scale batch processing via the Forge API, this skill provides the necessary guidance and implementation patterns for cutting-edge protein engineering and bioinformatics workflows.

Características Principales

81 GitHub stars
Function-conditioned generation and automated protein annotation
Multimodal protein design across sequence, structure, and function tracks
Inverse folding for sequence design based on target 3D structures
High-quality protein embeddings using ESM C models
Support for both local model execution and cloud-based Forge API integration

Casos de Uso

Generating protein embeddings for downstream machine learning tasks and similarity analysis
Designing novel proteins with specific functional properties or structural constraints
Conducting high-throughput protein engineering using the Forge API for scalable inference

Acerca de

Características Principales

81 GitHub stars
Function-conditioned generation and automated protein annotation
Multimodal protein design across sequence, structure, and function tracks
Inverse folding for sequence design based on target 3D structures
High-quality protein embeddings using ESM C models
Support for both local model execution and cloud-based Forge API integration

Casos de Uso

Generating protein embeddings for downstream machine learning tasks and similarity analysis
Designing novel proteins with specific functional properties or structural constraints
Conducting high-throughput protein engineering using the Forge API for scalable inference