BioSymphony
Skills, tool references, and command-line tools that let coding agents do biological research.
Six toolkits
Pick the toolkit that fits your question
BioProspector starts with a target molecule and possible routes. GeneCluster starts with candidate genes and genome or transcriptome evidence. Structure Factory covers prediction and design; CryoCore covers experimental maps and model validation.
BioProspector
Compare biosynthetic routes by evidence, gene count, host fit, and unresolved steps. Outputs Route comparisons, ranked enzyme and gene candidates, follow-up plans.
GeneCluster
Find candidate biosynthetic genes across species and compare their genomic neighborhoods. Outputs Candidate-gene tables, genome-context comparisons, reports with sources and evidence gaps.
Structure Factory
Plan binder design, structure prediction, and screening with comparable inputs and scoring across methods. Outputs Validated stage plans, compute requirements, task packs, candidate-ranking schemas.
CryoCore
Organize cryo-EM map and model validation, conformational-state comparison, and structural figures. Outputs A local HTML review demo, validation plans, figure specifications.
Small Molecules
Choose methods for molecular generation, synthesis planning, docking, affinity estimation, and property prediction. Outputs Tool comparisons, setup requirements, separate code, model, and data license records.
Ferm DoE
Generate fermentation experiment designs, analyze supplied results, and plan the next batch within declared constraints. Outputs Design tables, measurement-readiness reports, follow-up batches, scale-transfer reviews.Get started
Open a toolkit in your agent
- Clone one repo and open it in Codex, Claude Code, or another coding agent.
- Paste the first request. The agent reads the toolkit's instructions and runs its local example.
- Review the outputs, evidence gaps, and dependencies before configuring larger runs.
You set the research question and interpret the evidence. Planning outputs and computational rankings need experimental validation before they support biological claims.
Read this toolkit's README and agent instructions. Use its local example or worked reference to show what it produces. Keep this first pass local. Explain the output files, evidence gaps, and dependencies for my question.
Companion tools
Adjacent projects
Symphony NeoCloud
Run agent workflows on cloud compute with preflight checks, cost caps, artifact receipts, and cleanup.
Proteus
Structural biology automation for AI coding agents across PyMOL, ChimeraX, AlphaFold, PDB, UniProt, and Rosetta.
BioVoice
Voice control for PyMOL, ChimeraX, AlphaFold, and Rosetta on the OpenAI Realtime API.