[Integrations](https://spiral.ai/integrations) › [Muse Code](https://spiral.ai/integrations/muse-code) › [AlphaFold missense predictions](https://spiral.ai/integrations/muse-code/alphafold-missense-predictions)

# How to Use AlphaFold missense predictions in Muse Code

Get AlphaMissense pathogenicity scores and categories for amino acid substitutions in canonical human proteins. Use them to compare candidate changes and prioritize protein experiments.

|  |  |
| --- | --- |
| Model | AlphaFold missense predictions |
| Available through | Alexandria |
| Category | Health |
| Price | $0.00873 per call. |
| Returns | json |
| Runs | In the background |
| Setup | About 2 minutes, no API key |

## Prompts to try with AlphaFold missense predictions in Muse Code

- Use AlphaFold missense predictions for human TP53 changes R175H, R248Q, and R273H, then add their returned scores and categories to a CSV in this project.
- Use AlphaFold missense predictions for human EGFR changes L858R, T790M, and C797S, then build a sortable comparison table in the project's research dashboard.
- Use AlphaFold missense predictions for human CFTR changes G551D, G542S, and R117H, then create a notebook plot comparing their returned scores.
- Use AlphaFold missense predictions for human BRCA1 changes C61G and M1775R, then join the results to matching rows in the project's public assay CSV.
- Use AlphaFold missense predictions for human HBB changes E7V and E7K, then add a worked example to the project's protein genetics tutorial using the returned scores and categories.

## How to set up AlphaFold missense predictions in Muse Code

Pick the way that suits you. Each takes about two minutes.

### Use Spiral's MCP server

1. **Install Muse Code**

```bash
curl -fsSL https://dev.meta.ai/install.sh | bash
```

2. **Ask Muse to add the server**: Start `muse` and ask:

```text
Add the Spiral MCP server at https://api.spiral.ai/mcp
```

3. **Sign in**: In your terminal, run:

```bash
muse mcp login spiral
```

### Use the Spiral skill

1. **Send Muse Code this message**: Your agent reads Spiral's skill and sets itself up.

```
set up https://spiral.ai/SKILL.md
```

2. **Sign in or paste a key**: Approve Spiral in the browser, or give the agent a key from the console.

### Use the Spiral CLI

1. **Install and sign in**: Installs the Spiral CLI, signs you in and adds its skill to your coding agents.

```bash
npx -y spiral-cli@latest init
```

### Call the Spiral API

1. **Create an API key**: In the console, under API keys.

2. **Run AlphaFold missense predictions**: One request; the result comes back by polling or a webhook.

```bash
curl https://api.spiral.ai/v1/runs \
  -H "Authorization: Bearer $SPIRAL_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $(uuidgen)" \
  -d '{
    "provider": "alexandria",
    "model": "alphafold-ebi-ac-uk/variants/missense",
    "input": {
      "id": "<id>"
    }
  }'
```

Then paste one of the prompts above: Muse Code finds AlphaFold missense predictions on Spiral, shows the price and runs it.

Setup guide in the docs: https://spiral.ai/docs/quickstarts/muse-code

## AlphaFold missense predictions tools and prices

- [AlphaFold Protein Structure Database · Missense](https://spiral.ai/tools/alexandria/alphafold-ebi-ac-uk/variants/missense) (Alexandria): $0.00873 per call.

## Settings you can ask for

| Setting | Allowed values | Default | Description |
| --- | --- | --- | --- |
| id |  | — | UniProt accession (P04637), isoform (P04637-2), entry name (P53_HUMAN), AlphaFold DB model ID (AF-P04637-F1, AF-0000000000710001) or alphafold.ebi.ac.uk entry URL; case-insensitive. Exactly one of id and gene. |
| gene |  | — | Gene symbol (TP53, CD52, HLA-A), resolved to the one reviewed (Swiss-Prot) UniProt entry with that gene name, or that synonym, in organism. A symbol shared by several entries is invalid_input listing them. |
| organism |  | — | Organism for gene: NCBI taxonomy ID (9606) or name (human, Mus musculus). Default 9606 (human). Only with gene. |
| variants |  | — | Protein-level missense changes on the canonical sequence: one-letter (R175H) or three-letter (p.Arg175His) form. |

## What can you do with AlphaFold missense predictions in Muse Code?

Add AlphaMissense scores and categories to a protein research table in your project. Your coding agent can join results to public assay measurements and build a comparison view.

### Choose substitutions for functional assays

Compare scores for candidate TP53 substitutions and select changes spanning the returned categories for experimental testing.

### Annotate a public research table

Add AlphaMissense scores and categories to published CFTR amino acid changes so researchers can sort the table by predicted pathogenicity.

### Compare changes at one residue

Compare alternative substitutions at TP53 residue 175 to see how the predicted score depends on the replacement amino acid.

### Review uncertain predictions

Separate uncertain EGFR predictions from likely benign and likely pathogenic results when preparing a research review.

### Compare predictions with assay results

Join returned BRCA1 scores to a public functional assay table and identify disagreements worth examining.

### Prepare a protein genetics lesson

Use a small set of HBB substitutions to teach the difference between a numerical pathogenicity prediction and its assigned category.

## Overview

- [Alexandria](https://spiral.ai/providers/alexandria)
- [All Muse Code guides](https://spiral.ai/integrations/muse-code)

## AlphaFold missense predictions in other agents

- [Amp](https://spiral.ai/integrations/amp/alphafold-missense-predictions)
- [ChatGPT](https://spiral.ai/integrations/chatgpt/alphafold-missense-predictions)
- [Claude](https://spiral.ai/integrations/claude-desktop/alphafold-missense-predictions)
- [Claude Code](https://spiral.ai/integrations/claude-code/alphafold-missense-predictions)
- [Cline](https://spiral.ai/integrations/cline/alphafold-missense-predictions)
- [Codex](https://spiral.ai/integrations/codex/alphafold-missense-predictions)
- [Cursor](https://spiral.ai/integrations/cursor/alphafold-missense-predictions)
- [Devin](https://spiral.ai/integrations/devin/alphafold-missense-predictions)
- [dots](https://spiral.ai/integrations/dots/alphafold-missense-predictions)
- [Factory Droid](https://spiral.ai/integrations/factory-droid/alphafold-missense-predictions)
- [Gemini](https://spiral.ai/integrations/gemini/alphafold-missense-predictions)
- [Gemini CLI](https://spiral.ai/integrations/gemini-cli/alphafold-missense-predictions)
- [Google Antigravity](https://spiral.ai/integrations/antigravity/alphafold-missense-predictions)
- [goose](https://spiral.ai/integrations/goose/alphafold-missense-predictions)
- [Grok](https://spiral.ai/integrations/grok/alphafold-missense-predictions)
- [Grok Build](https://spiral.ai/integrations/grok-build/alphafold-missense-predictions)
- [Hermes](https://spiral.ai/integrations/hermes/alphafold-missense-predictions)
- [Instinct](https://spiral.ai/integrations/instinct/alphafold-missense-predictions)
- [Manus](https://spiral.ai/integrations/manus/alphafold-missense-predictions)
- [Muse](https://spiral.ai/integrations/muse/alphafold-missense-predictions)
- [n8n](https://spiral.ai/integrations/n8n/alphafold-missense-predictions)
- [Notion](https://spiral.ai/integrations/notion/alphafold-missense-predictions)
- [OpenClaw](https://spiral.ai/integrations/openclaw/alphafold-missense-predictions)
- [opencode](https://spiral.ai/integrations/opencode/alphafold-missense-predictions)
- [Perplexity](https://spiral.ai/integrations/perplexity/alphafold-missense-predictions)
- [Raycast](https://spiral.ai/integrations/raycast/alphafold-missense-predictions)
- [VS Code](https://spiral.ai/integrations/vscode/alphafold-missense-predictions)
- [Windsurf](https://spiral.ai/integrations/windsurf/alphafold-missense-predictions)
- [Zapier](https://spiral.ai/integrations/zapier/alphafold-missense-predictions)

## More Muse Code guides

- [mempool.space Transaction output spending status](https://spiral.ai/integrations/muse-code/mempool-space-transaction-output-spending-status)
- [AlphaFold predicted protein complexes](https://spiral.ai/integrations/muse-code/alphafold-predicted-protein-complexes)
- [mempool.space Recent Bitcoin blocks](https://spiral.ai/integrations/muse-code/mempool-space-recent-bitcoin-blocks)
- [AlphaFold protein structure prediction](https://spiral.ai/integrations/muse-code/alphafold-protein-structure-prediction)
- [mempool.space Block transactions](https://spiral.ai/integrations/muse-code/mempool-space-block-transactions)
- [BDPM HAS medicine assessments](https://spiral.ai/integrations/muse-code/bdpm-has-medicine-assessments)

## Questions about AlphaFold missense predictions in Muse Code

**Can Muse Code use AlphaFold missense predictions?**

Yes. Add Spiral to Muse Code as an MCP server, and Muse Code can find and run AlphaFold missense predictions like any built-in tool.

**How much does AlphaFold missense predictions cost in Muse Code?**

$0.00873 per call. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Muse Code plan is separate.

**Do I need an Alexandria account to use AlphaFold missense predictions in Muse Code?**

No. Spiral runs AlphaFold missense predictions on its own Alexandria account. You sign in to Spiral once and never paste a key.

**How long does it take to set up AlphaFold missense predictions in Muse Code?**

About two minutes: install Muse Code, then ask Muse to add the server, then sign in.

**Does Muse Code wait while AlphaFold missense predictions runs?**

AlphaFold missense predictions runs in the background. Muse Code starts the run, checks on it and gives you the json when it's ready.

**Which proteins have AlphaMissense data in AlphaFold DB?**

AlphaFold DB presents AlphaMissense annotations for canonical human proteins. The underlying canonical substitution dataset covers about 20,000 UniProt proteins and 216 million amino acid changes.

**What does the pathogenicity score mean?**

Scores run from 0 to 1. EMBL-EBI describes three ranges: 0 to 0.34 as likely benign, 0.34 to 0.564 as uncertain, and 0.564 to 1 as likely pathogenic.

**How accurate are the classifications?**

DeepMind reports that its published threshold classified 89% of missense changes with 90% precision against a database of known disease annotations. That benchmark does not establish the accuracy of every individual prediction.

**How current are the protein references?**

The underlying dataset uses UniProt release 2021_02 for protein identifiers. It supplies precomputed predictions, so check sequence compatibility before applying its scores to a newer protein reference.

**Can these predictions support a clinical diagnosis?**

EMBL-EBI states that AlphaFold and its outputs have not been validated or approved for clinical use. Use the scores for research comparisons and experiment planning.

