Seq Design BuddyGitHub

Beginner guide

From benchmark planning to a visual genetic design

Follow the handoffs in order when you need the full workflow. Each stage is a separate tool with its own setup, files, and execution environment.

  1. 01

    SeqTrainer BenchLab

    Plan and benchmark

    Input

    Labeled CSV or TSV

    Activities

    Inspect columns, class balance, preprocessing, and verified lightweight baselines.

    Output

    run_config.json, benchmark_plan.json, manifests, and optional metrics/predictions.

  2. 02

    SeqTrainer

    Annotate a plasmid

    Input

    GenBank plasmid, checkpoint, and benchmark manifest

    Activities

    Scan sequence windows, score possible promoters, preserve existing features, and add predicted features.

    Output

    Annotated .gb, predictions.csv, and annotation_manifest.json.

  3. 03

    SBOL Validator

    Validate the design

    Input

    SBOL .nt/.rdf, GenBank, FASTA, or GFF3 file

    Activities

    Validate only, or validate and convert using the official Validator.

    Output

    Validation report and optional converted file.

  4. 04

    SBOL Canvas

    Visualize the design

    Input

    SBOL2 RDF/XML .rdf file for the SeqTrainer Canvas handoff

    Activities

    Import through File -> Import, inspect glyphs, and edit or export in Canvas.

    Output

    A visual genetic design.

Choose your path

Path A: Benchmark planning only

You want a clean benchmark plan and reproducibility bundle, not annotations.

Start: Start with BenchLab.

Success: Stop after downloading run_config.json, benchmark_plan.json, and manifests.

Path B: Existing checkpoint annotation

You already have a GenBank plasmid, checkpoint, and matching benchmark manifest.

Start: Start with SeqTrainer annotation.

Success: Stop after annotated .gb, predictions.csv, and annotation_manifest.json.

Path C: Annotation with SBOL export

You need machine-readable SBOL3 plus an optional Canvas compatibility file.

Start: Use the SeqTrainer SBOL export branch.

Success: Validate annotated.nt and carry annotated_sbol2.rdf to Canvas when needed.

Path D: Visualize existing SBOL design

You already have an SBOL design and only need checking plus visualization.

Start: Start with SBOL Validator.

Success: Import the validated SBOL2 .rdf file into SBOL Canvas through File -> Import.

Static file handoff table

These files move between the original tools. Seq Design Buddy only documents the handoff and never receives the files.

Source toolOutput fileMeaningNext toolImportant limitation
BenchLabrun_config.jsonSettings and dataset identity for reproducing a benchmark.External training or sharing the plan.It does not include raw dataset rows.
BenchLabmetrics.json / predictions.csvSmall local baseline results when a baseline actually ran.Review against later model results.Can be empty for plan-only exports.
SeqTrainer annotation-mvpannotated .gbGenBank plasmid with preserved source features and predicted promoter features.Review in a GenBank-aware viewer or keep as primary visual annotation output.Dummy mode is a file-writing smoke test only.
SeqTrainer annotation-mvppredictions.csv / annotation_manifest.jsonScored windows plus provenance, settings, warnings, and threshold source.Audit the prediction run.Checkpoint and manifest must match the same completed benchmark run.
SeqTrainer annotation-sbol3-labeled-promotersannotated.ntCanonical SBOL3 N-Triples machine-exchange output.Validate with SBOL Validator.Do not upload this file to Canvas for the documented Canvas workflow.
SeqTrainer annotation-sbol3-labeled-promotersannotated_sbol2.rdfSBOL2 RDF/XML compatibility output for Canvas glyph import.Import into SBOL Canvas.It is compatibility output; inspect the Canvas layout manually.