How to Use AlphaFold predicted protein complexes in Devin
Find predicted protein complexes containing your protein, with chain composition and interface confidence scores. Use these results to shortlist interaction hypotheses for research and compare assemblies beyond a single protein's fold.
Prompts to try with AlphaFold predicted protein complexes in Devin
Copy an example into Devin, or ask in your own words.
Use AlphaFold predicted protein complexes for human TP53. Add the returned complex identifiers, chain compositions, and ipTM scores to a CSV in this project.
Use AlphaFold predicted protein complexes for P69905. Build a table in this project showing the proteins and copy counts in each returned assembly.
Use AlphaFold predicted protein complexes for human EGFR. Create a small report that orders the returned complexes by ipTM and highlights scores above 0.8.
Use AlphaFold predicted protein complexes for human BRCA1 and BARD1 separately. Compare the returned complex identifiers and save any shared predictions in this project's research notes.
Use AlphaFold predicted protein complexes for human CD274. Convert the returned protein compositions into a graph data file with complex identifiers and interface scores attached.
How to set up AlphaFold predicted protein complexes in Devin
Pick the way that suits you. Each takes about two minutes.
Use Spiral's MCP server
- Add a custom MCP
Open Customize › MCPs. Click + and choose New MCP. Fill in:
- Sign in to Spiral and approve
In Customize › MCPs, open Spiral and choose Connect. Spiral opens: sign in, pick the organization and click Authorize. No API key.
- Check it
On Spiral's configuration page, click Test tools. Devin lists Spiral's tools.
Use the Spiral skill
- Send Devin this message
Your agent reads Spiral's skill and sets itself up.
set up https://spiral.ai/SKILL.md
- Sign in or paste a key
Approve Spiral in the browser, or give the agent a key from the console.
Use the Spiral CLI
- Install and sign in
Installs the Spiral CLI, signs you in and adds its skill to your coding agents.
npx -y spiral-cli@latest init
Call the Spiral API
- Create an API key
In the console, under API keys.
- Run AlphaFold predicted protein complexes
One request; the result comes back by polling or a webhook.
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/interactions/complexes", "input": { "id": "<id>" } }'
Then paste one of the prompts above: Devin finds AlphaFold predicted protein complexes on Spiral, shows the price and runs it.
AlphaFold predicted protein complexes providers and prices
Spiral runs AlphaFold predicted protein complexes through Alexandria. Open a card for today's price, with every fee included.
Settings you can ask for
Ask Devin for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Limit | 1 to 100 | 25 | Most complexes to return, highest ipTM first. |
| Min iptm | 0 to 1 | Only complexes with ipTM at least this (0.8 keeps confident interfaces). |
Every API field of AlphaFold Protein Structure Database · Complexes is on its tool page: open its card above.
What can you do with AlphaFold predicted protein complexes in Devin?
Add predicted protein assemblies and interface confidence to a target table in your project. A coding agent can turn the returned chain compositions into tables or graphs that link each prediction to its complex identifier.
Prioritize interaction experiments
Retrieve complexes containing human TP53 and rank the returned assemblies by interface confidence to prepare a shortlist for experimental review.
Compare predicted chain composition
Inspect the proteins and copy counts in returned assemblies to distinguish predicted self-association from complexes involving different proteins.
Annotate a research target list
Add available complex identifiers, chain composition, and interface scores to a list of protein targets before choosing structures for closer review.
Build a predicted interaction graph
Turn returned protein compositions into graph connections, retaining complex identifiers and confidence scores so researchers can inspect the underlying predictions.
Watch for changes in available predictions
Repeat the same protein lookups monthly and compare saved results to identify newly returned complexes or changed interface scores.
Prepare a structural biology lesson
Use returned assemblies to show how chain composition and interface confidence answer different questions about a predicted protein complex.
More on AlphaFold predicted protein complexes and Devin
Overview
AlphaFold predicted protein complexes in other agents
Questions about AlphaFold predicted protein complexes in Devin
Can Devin use AlphaFold predicted protein complexes?
Yes. Add Spiral to Devin as an MCP server, and Devin can find and run AlphaFold predicted protein complexes like any built-in tool.
How much does AlphaFold predicted protein complexes cost in Devin?
$0.00873 per call. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Devin plan is separate.
Do I need an Alexandria account to use AlphaFold predicted protein complexes in Devin?
No. Spiral runs AlphaFold predicted protein complexes 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 predicted protein complexes in Devin?
About two minutes: add a custom MCP, then sign in to Spiral and approve, then check it.
Does Devin wait while AlphaFold predicted protein complexes runs?
AlphaFold predicted protein complexes runs in the background. Devin starts the run, checks on it and gives you the json when it's ready.
Does AlphaFold include complexes of identical and different proteins?
Yes. AlphaFold DB includes homodimers, which contain two identical proteins, and heterodimers, which contain two different proteins.
How recently has AlphaFold expanded its complex coverage?
EMBL-EBI announced a viral complex expansion on September 24, 2026, providing predictions for over 2,800 viruses. That is a documented dataset addition, not a promise of daily updates.
Are the complex predictions on the AlphaFold website already filtered?
Yes. Displayed complex entries pass thresholds of pDockQ2max ≥ 0.23 and ipSAEmax ≥ 0.6, so their score distribution excludes predictions below those cutoffs.
What does an ipTM score tell me?
ipTM measures confidence in how the protein chains fit together. EMBL-EBI describes scores above 0.8 as confident predictions and scores below 0.6 as suggesting a failed prediction.
Why distinguish pLDDT from interface confidence?
pLDDT measures local confidence residue by residue. For a complex, ipTM addresses a different question: confidence in the relative positions of its chains.
Start using AlphaFold predicted protein complexes in Devin
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Devin:
set up https://spiral.ai/SKILL.md


