ESM Protein Design & Modeling

关于

The ESM skill integrates state-of-the-art protein language models into Claude, facilitating complex tasks across sequence, structure, and function tracks. By leveraging ESM3 for generative multimodal protein design and ESM C for high-efficiency representation learning, users can perform inverse folding, predict 3D structures, and design novel proteins with specific functional annotations. This toolkit supports both local model execution for prototyping and the cloud-based Forge API for high-throughput, production-grade inference, making it an essential resource for bioinformaticians and machine learning researchers.

主要功能

• Function-conditioned generation and annotation
• High-quality protein embeddings for downstream ML tasks
• 81 GitHub stars
• Automated structure prediction and inverse folding
• Asynchronous batch processing via the Forge API
• Multimodal generative protein design across sequence and structure

使用场景

• Generating protein embeddings for functional classification or similarity analysis
• Designing novel fluorescent proteins or enzymes with specific functions
• Predicting 3D protein coordinates from amino acid sequences