skills.rest
junior1p
junior1pCOMMUNITY
Explore all skills by this author →
GitHub

protein-design-workflow

Orchestrate protein binder design workflows from target preparation to candidate filtering.

11|Updated Mar 4, 2026
Search Tags:#workflow#pipeline#chai#protein-design#binder-design#structure-validation#boltzgen
One-click install
npx skills add https://github.com/junior1p/ProteinClaw --skill protein-design-workflow-junior1p
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: protein-design-workflow
Source: https://github.com/junior1p/ProteinClaw/tree/main/skills/protein-design-workflow
Command: npx skills add https://github.com/junior1p/ProteinClaw --skill protein-design-workflow-junior1p

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides structured, end-to-end guidance to design and validate protein binders, removing guesswork across target preparation, generative design, validation, and filtering so researchers can convert a structural target into prioritized candidate binders.

Core Features & Use Cases

  • Target preparation & hotspot selection: instructions for fetching PDB entries, extracting chains, trimming to binding regions, and selecting exposed hotspot residues.
  • Generative design orchestration: guidance on running BoltzGen or BindCraft to produce all-atom designs with configurable counts and protocols.
  • Structure validation & filtering: recommended use of Chai-1 or Boltz for predictions, collection of metrics (pLDDT, ipTM, PAE), and example filtering and scoring strategies to rank candidates.
  • Resource planning & checkpoints: GPU/time/cost estimates and quality checkpoints to ensure campaign efficiency and result reliability.
  • Use Case: Run a 100-design campaign for a given target, predict structures, score with pLDDT/ipTM/PAE, and select the top 50 candidate binders.

Quick Start

Use the protein-design-workflow skill to prepare the target, generate 100 BoltzGen designs, validate with Chai-1 predictions, and filter the top 50 candidates by pLDDT and ipTM.

Frequently Asked Questions about protein-design-workflow

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

FAQPage Schema
How do I run an end-to-end protein binder design workflow from target preparation to candidate filtering?▼

To execute protein binder design, fetch PDB or UniProt structural targets, run generative design with BoltzGen or BindCraft, validate structures using Chai-1 or Boltz, and rank candidates by pLDDT, ipTM, and PAE metrics.

What metrics are used to filter and score protein binder candidates?▼

Protein binder candidates are filtered and scored using pLDDT, ipTM, and PAE metrics collected from Chai-1 or Boltz structure predictions to rank and prioritize designs.

Can I use BoltzGen and BindCraft for generative protein design in the same campaign?▼

The workflow supports both BoltzGen and BindCraft for generative protein design, allowing configurable design counts and protocols to produce all-atom binder candidates within a single campaign.

Do I need GPU compute for structure validation and generative design steps?▼

GPU compute is required for generative design and structure prediction steps, with the workflow providing GPU time and cost estimates to ensure campaign efficiency and result reliability.

How do I prepare a PDB target for protein binder design?▼

Target preparation involves fetching PDB entries, extracting relevant chains, trimming to binding regions, and selecting exposed hotspot residues to ready the target for generative binder design.

What is the best way to plan a 100-design protein binder campaign?▼

A 100-design protein binder campaign requires estimating GPU time and costs, configuring BoltzGen or BindCraft protocols, setting quality checkpoints, and filtering the top 50 candidates by pLDDT and ipTM scores.

Related Skills

Explore More Education & Research Skills→
NVIDIA-BioNeMoNVIDIA-BioNeMo
413

complexa-target

Register and validate Proteina-Complexa design targets in YAML configuration files.

Official
Advanced
alex-hhalex-hh

pgdh_boltz2

Cross-validate 15-PGDH binder designs using Boltz-2 structure prediction.

Community
Advanced
junior1pjunior1p
11

setup

Automate Modal CLI installation, authentication, and biomodals repository bootstrap for protein design toolchains.

Community
Intermediate
junior1pjunior1p
11

protein-qc

Filter and rank protein designs using pLDDT, ipTM, PAE, and sequence liability thresholds.

Community
Advanced