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.
Key Features
01Function-conditioned generation and annotation
02High-quality protein embeddings for downstream ML tasks
0381 GitHub stars
04Automated structure prediction and inverse folding
05Asynchronous batch processing via the Forge API
06Multimodal generative protein design across sequence and structure
