chai

Predict protein structures from FASTA sequences using Chai-1.

151|20|Updated Jan 15, 2026
One-click install
npx skills add https://github.com/adaptyvbio/protein-design-skills --skill chai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: chai
Source: https://github.com/adaptyvbio/protein-design-skills/tree/main/skills/chai
Command: npx skills add https://github.com/adaptyvbio/protein-design-skills --skill chai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill enables rapid, scalable prediction of protein structures from sequence data using Chai-1, accelerating design, validation, and discovery in computational biology.

Core Features & Use Cases

  • Structure prediction: Generate high-quality protein-protein and protein-ligand complex structures from sequences.
  • Validation workflows: Validate designed binders and interfaces with rapid, scalable predictions.
  • High-throughput API access: Use the Chai API for batch predictions and integration into pipelines.
  • Use Case: When screening dozens to hundreds of designs, run automated structure predictions and compare scores to select top candidates.

Quick Start

Run a quick prediction on a provided FASTA using the biomodals workflow: chai-lab predict --fasta complex.fasta --output predictions/

Frequently Asked Questions about chai

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

FAQPage Schema
How do I predict protein-protein complex structures from FASTA sequences?

You can predict protein-protein complex structures by passing FASTA sequences to the Chai-1 workflow, which outputs CIF structure files and evaluation metrics like plddt and pae.

Can I run high-throughput protein structure predictions via an API?

Yes, high-throughput protein structure prediction is supported via the Chai API, allowing you to batch process sequences and integrate predictions directly into screening pipelines.

What parameters can I tune for protein-ligand complex structure prediction?

When predicting protein-ligand complex structures, you can adjust parameters including num_trunk_recycles, num_diffn_timesteps, and seed to refine the resulting CIF outputs.

Does Chai-1 support binder validation workflows for designed proteins?

Yes, Chai-1 supports binder validation by generating rapid, scalable structure predictions, allowing you to compare scoring metrics and select top design candidates.

What output formats and metrics are generated when predicting protein structures?

Predicting protein structures generates CIF structure files alongside metrics such as plddt and pae, providing quantitative confidence scores for your predicted complexes.