skills.rest
openai
openaiOFFICIAL
Explore all skills by this author →
WebsiteGitHub

boltz-protein-screen

Screen and rank protein binder candidates against a target using the Boltz API.

5.3k|765|Updated Mar 4, 2026
Search Tags:#api#boltz#antibody#structural biology#binding confidence#protein screening#binder ranking
One-click install
npx skills add https://github.com/openai/plugins --skill boltz-protein-screen
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: boltz-protein-screen
Source: https://github.com/openai/plugins/tree/main/plugins/boltz-api-cli/skills/boltz-protein-screen
Command: npx skills add https://github.com/openai/plugins --skill boltz-protein-screen

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Ranking a library of candidate protein binders against a target requires submitting structured payloads to the Boltz API, estimating costs, managing long-running jobs, and interpreting confidence metrics. This Skill automates that entire screening workflow so you can go from a binder library to ranked hits without manual API orchestration.

Core Features & Use Cases

  • Library Screening: Submit protein, peptide, antibody, or nanobody candidate libraries against a target using either a structure template (CIF/PDB) or sequence-only input.
  • Cost Estimation & Confirmation: Run estimate-cost before submitting so you always see the exact USD cost and confirm spend before starting a job.
  • Background Download & Ranking: Download results in the background and rank hits by binding_confidence with iptm and min_interaction_pae as tiebreakers.
  • Use Case: You have 200 candidate nanobody sequences and a target CIF file. Use this Skill to build the payload, confirm the cost, submit the screen, and receive a ranked list of the top binders with predicted structures.

Quick Start

Use the boltz-protein-screen skill to estimate the cost and rank these candidate binder sequences against my target structure file.

Frequently Asked Questions about boltz-protein-screen

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

FAQPage Schema
How do I screen a protein binder library with Boltz?▼

Build a payload with a proteins list of candidate complexes and a target variant, run boltz-api protein:library-screen estimate-cost, confirm the cost, then submit with start. Download results in the background and rank hits from results/index.jsonl by binding_confidence.

What target formats does Boltz library screening support?▼

Two variants are supported: structure_template for CIF/PDB files or URLs with chain selection and residue cropping, and no_template for sequence-only targets with optional epitope residue hints. All residue indices are 0-based.

Can Boltz protein screen design new proteins or screen small molecules?▼

No. This endpoint only ranks existing protein, peptide, antibody, or nanobody candidates against a target. It is not intended for de novo protein design or small-molecule screening.

How is Boltz library screening cost calculated?▼

Cost is tiered by total complex length (target plus candidate), so there is no flat per-candidate rate. Always run estimate-cost on the full payload and quote only the USD figure it returns.

Why is my Boltz screening result missing optimization_score?▼

The protein:library-screen endpoint does not emit optimization_score, so sorting by it returns an empty list. Rank by binding_confidence descending, using iptm and min_interaction_pae as tiebreakers.

Related Skills

Explore More Education & Research Skills→
junior1pjunior1p
11

iggm

Design antigen-conditioned antibody and nanobody CDR loops from X-masked FASTA inputs.

Community
Intermediate
LiorZLiorZ
2

boltz

Predict biomolecular structures and affinities from YAML inputs via CLI diffusion workflow.

Community
Advanced
NVIDIA-BioNeMoNVIDIA-BioNeMo
413

alphafold-database-fetch-and-analyze

Fetch AlphaFold protein structures and analyze pLDDT and PAE metrics.

Official
Intermediate
K-Dense-AIK-Dense-AI
41.1k

tamarind

Submit molecular design and structure prediction jobs to the Tamarind Bio cloud API.

Official
Advanced