chai

Predict protein structures with Chai-1 and output CIF models with scores.

11|Updated Mar 4, 2026
One-click install
npx skills add https://github.com/junior1p/ProteinClaw --skill chai-junior1p
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: chai
Source: https://github.com/junior1p/ProteinClaw/tree/main/skills/chai
Command: npx skills add https://github.com/junior1p/ProteinClaw --skill chai-junior1p

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Chai-1 enables rapid, automated protein structure predictions to accelerate design validation, reducing manual modeling time and errors when evaluating complexes and ligand interactions.

Core Features & Use Cases

  • Automated structure prediction for protein complexes and protein-ligand systems using the Chai-1 model.
  • High-throughput predictions via the Chai API for batch design campaigns.
  • Output in CIF format with per-model scores (pLDDT, ipTM, PAE) and support for downstream QC and comparison with AlphaFold2.

Quick Start

Provide a FASTA file containing the target sequence and run the Chai-1 API to generate predicted structures and confidence scores.

Frequently Asked Questions about chai

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

FAQPage Schema
How do I predict protein structures for protein-ligand complexes?

Predict protein structures for protein-ligand complexes by providing a FASTA file to the Chai-1 model, which infers high-accuracy structures and outputs CIF models along with confidence scores.

What confidence scores does Chai-1 output for protein-protein complex prediction?

Protein-protein complex prediction outputs CIF models accompanied by per-model scores including pLDDT, ipTM, and PAE to support downstream quality control and comparison.

Do I need a CUDA 12.x compatible GPU to run Chai-1 molecular modeling?

Running Chai-1 molecular modeling requires Python 3.10+, CUDA 12.x, and a compatible GPU to execute the chai_lab API for generating predicted structures.

Can I run high-throughput protein structure predictions in batch design campaigns?

High-throughput protein structure predictions are supported via the Chai API, enabling automated batch design campaigns to evaluate protein complexes and ligand interactions rapidly.

What is the best way to automate design validation for protein complexes?

Automate design validation for protein complexes using the Chai-1 foundation model to reduce manual modeling time and errors when evaluating complexes and ligand interactions.

How does Chai-1 structure prediction compare to AlphaFold2 for downstream QC?

Chai-1 structure prediction outputs CIF models with pLDDT, ipTM, and PAE scores, explicitly supporting downstream QC and comparison with AlphaFold2 results.