SeqTrainer

SeqTrainer is a synthetic biology ML domain toolkit focused on connecting SBOL / SynBioHub data to modern model workflows.

It is designed to be complementary to Keras and PyTorch rather than replacing them.

What this refactor introduces

• Clear package layering under seqtrainer/
• Framework-neutral core (clients, sparql, data, transforms, models)
• Optional framework adapters (seqtrainer.keras, seqtrainer.torch)
• Graph-focused utilities in seqtrainer.graph
• Application-level API entrypoints in seqtrainer.applications
• CLI foundation (seqtrainer command)

Install

pip install -e .

Optional extras:

pip install -e '.[torch]'
pip install -e '.[keras]'
pip install -e '.[gnn]'
pip install -e '.[dev]'

Package layout

• seqtrainer/clients: SynBioHub and remote clients (auth/retry/pagination-ready client)
• seqtrainer/sparql: prefixes, builders, canonical query recipes, and typed result normalization
• seqtrainer/data: SBOL loaders, recipes, materialized datasets, local snapshots/cache manifests
• seqtrainer/transforms: DNA transforms and feature extraction
• seqtrainer/models: framework-neutral backbone/head registries and default model specs
• seqtrainer/keras: Keras model factories + tensorized tf.data adapters from MaterializedDataset (optional dependency)
• seqtrainer/torch: PyTorch tensorized adapters/dataloaders, HF/DNABERT backbones, and regression/classification heads (optional dependency)
• seqtrainer/graph: RDF/SBOL graph conversion, schema extraction, and graph-config builders
• seqtrainer/applications: task-oriented blueprints
• seqtrainer/cli: command-line entrypoints

CLI examples

seqtrainer sparql prefixes
seqtrainer inspect-sbol data/sbol_data/sample_design_0.xml
seqtrainer dataset recipes
seqtrainer dataset build data/sbol_data/sample_design_0.xml --cache
seqtrainer model build --framework torch --task regression

Status

This is the first architecture-focused cleanup. Some framework integrations are intentionally placeholders with TODOs to keep a stable, minimal public surface.

CI checks

The repository includes a GitHub Actions CI matrix for:

• base test suite
• torch extra
• keras extra
• gnn extra
• lint (ruff) and type checks (mypy)
• packaging build + twine check