rfd3-reference

Reference RFdiffusion3 parameters, contig syntax, and design scenarios.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/polaires/Banta_Lab_RFdiffusion --skill rfd3-reference
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: rfd3-reference
Source: https://github.com/polaires/Banta_Lab_RFdiffusion/tree/main/.claude/skills/rfd3-reference
Command: npx skills add https://github.com/polaires/Banta_Lab_RFdiffusion --skill rfd3-reference

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill provides comprehensive guidance and reference material for using RFdiffusion3 for protein design, helping users understand and apply its advanced features.

Core Features & Use Cases

  • Parameter Reference: Detailed explanations of key RFdiffusion3 input parameters like contig, select_fixed_atoms, select_hotspots, and symmetry options.
  • Design Type Guidance: Instructions and examples for various design scenarios including small molecule binders, protein binders, enzymes, symmetric oligomers, and metal-binding proteins.
  • Best Practices: Essential tips for successful protein design, such as optimal batch sizes for symmetry and expected success rates for specific conditioning.

Quick Start

Use the rfd3-reference skill to understand how to specify a two-chain protein design with contig syntax "60-80,/0,60-80".

Frequently Asked Questions about rfd3-reference

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I use contig syntax for multi-chain protein design in RFdiffusion3?

RFdiffusion3 uses contig syntax to define chain lengths and breaks, such as "60-80,/0,60-80" for a two-chain protein design. This syntax maps directly to the generated structural topology.

What symmetry types are supported for symmetric oligomer design in RFdiffusion3?

RFdiffusion3 supports various symmetry types for designing symmetric oligomers. Best practices include using optimal batch sizes for symmetry generation to ensure successful computational protein design outputs.

How do I select fixed atoms and hotspots when designing protein binders?

Use the `select_fixed_atoms` and `select_hotspots` input parameters in RFdiffusion3 to constrain specific structural regions during protein binder design. These parameters anchor the design to desired interaction interfaces.

Does the RFdiffusion3 Foundry API support all CLI features from v1 and v2?

The RFdiffusion3 Foundry API has known limitations compared to the RFdiffusion v1/v2 CLI features. Users should review these constraints to ensure their specific protein design scenarios are fully supported.

What are the best practices for designing small molecule binders and enzymes with RFdiffusion3?

Best practices for small molecule binders and enzymes involve specific RFdiffusion3 conditioning parameters and understanding expected success rates. Proper parameter tuning is essential for functional computational protein design.

When should I avoid using RFdiffusion3 for protein design?

You should avoid RFdiffusion3 if your design requires unsupported CLI features absent from the Foundry API, or if your scenario exceeds the expected success rates for specific conditioning without extensive downstream validation.