Design, embed, and analyze proteins using ESM3 and ESM C.
npx skills add https://github.com/rubensliv/k-dense-ai --skill esm-rubensliv
Please help me install this Agent Skill. Skill: esm Source: https://github.com/rubensliv/k-dense-ai/tree/main/scientific-skills/esm Command: npx skills add https://github.com/rubensliv/k-dense-ai --skill esm-rubensliv
Protein researchers need integrated access to modern protein language models to design, embed, and evaluate proteins without managing multiple toolchains. This skill consolidates ESM3 multimodal design capabilities with ESM C embeddings, enabling end-to-end prototyping from sequence design to embedding-based analysis using either local models or Forge API.
Load an ESM3 model (local or Forge), create an ESMProtein with a partial sequence, and run a simple generation to complete the protein.
High-intent search queries and answers about installing and using this skill.
Compute protein embeddings for sequence similarity by loading an ESM C model and running it on your sequences to generate high-quality vectors for clustering and downstream ML tasks.
Design novel proteins with sequence and structure constraints by using ESM3 multimodal generation to create coordinated outputs across sequence, structure, and function annotation tracks.
Set up ESM3 for local inference or Forge API batch processing by installing the required esm package, then configuring the model load call to target either your local environment or the Forge endpoint.
Multimodal protein generation with ESM3 is the process of creating proteins by simultaneously generating coordinated outputs across sequence, structure, and function annotation tracks from partial inputs.
ESM3 supports partial sequence inputs for protein completion by allowing you to create an ESMProtein object with incomplete sequences and running a generation step to fill the missing residues.
Provide inputs including protein sequences, coordinates, and function annotations to run ESM3 embedding and analysis workflows, allowing the model to process multimodal data for generation and evaluation.
jackspaceDesign or optimize protein sequences, structures, and functions using ESM3 and ESM C models.
silversteinGenerate protein sequences, predict structures, and compute embeddings with ESM3 and ESM C.
SciMate-AIDesign proteins and generate embeddings using ESM3 and ESM C.
jimmc414Generate protein sequences and extract embeddings using ESM3 and ESM C.
