---
name: fair-esm2
description: Use ESM2 protein language models for embeddings, mutation scoring, remote homology, or representation analysis. Use when a task needs protein embeddings, zero-shot variant scores, clustering, or sequence-function triage.
---

# Fair ESM2

Use this skill for ESM2-style protein language model analysis.

Workflow:

1. Normalize proteins into FASTA with stable identifiers and mutation notation.
2. Verify the local package, checkpoint, endpoint, or notebook route before execution.
3. Save embeddings, variant-score tables, clustering plots, model version, command, and runtime logs.
4. Compare model-derived rankings against UniProt, structures, conservation, assays, or literature.
5. Mark sequence regions with poor coverage, disorder, low homology, or domain-boundary ambiguity.

Use ESM2 outputs as features or hypotheses unless a source-backed validation step supports the conclusion.