ProteinsGPU12 credits

RFdiffusion — Protein Backbone Generator

Generate novel protein backbones and enzyme scaffolds via denoising diffusion

RFdiffusion (Baker Lab, 2023) uses a denoising diffusion model to generate novel protein backbones, including binders for a specified target and enzyme active-site scaffolds. RFdiffusion2 (December 2025) extends this to functional enzyme design from active-site specifications. SciRouter exposes both via a unified API: paste a target PDB, specify hotspot residues, and receive ranked backbone designs with pLDDT confidence.

$0.12

per API call

credits per call

/v1/design/rfdiffusion

API endpoint

Features

Novel protein backbone generation via denoising diffusion

Target-conditioned binder design with hotspot residue support

Enzyme active-site scaffolding (RFdiffusion2)

Returns PDB + pLDDT + backbone RMSD per design

Ranked by confidence score

Async job pattern — submit + poll

Quick Start

RFdiffusion2 — Python Examplepython

import requests, time

API_KEY = "sk-sci-your-key-here"
url = "https://scirouter.ai/v1/design/rfdiffusion"

# Submit a design job — target PDB, hotspot residues, length range
resp = requests.post(
    url,
    json={
        "target_pdb": open("pdl1.pdb").read(),
        "hotspots": [42, 67, 89],  # PD-L1 interface residues
        "length_min": 65,
        "length_max": 85,
        "num_designs": 5,
    },
    headers={"Authorization": f"Bearer {API_KEY}"}
)
job = resp.json()["data"]
print(f"Job: {job['job_id']}, status: {job['status']}")

# Poll for results (usually returns quickly in mock mode)
while job["status"] == "processing":
    time.sleep(2)
    job = requests.get(f"{url}/{job['job_id']}",
        headers={"Authorization": f"Bearer {API_KEY}"}
    ).json()["data"]

for d in job["result"]["designs"][:3]:
    print(f"  {d['design_id']}: len={d['length']} pLDDT={d['plddt']}")