GRASP Library Designer

Codon-optimize GRASP (Farley et al., NAR 2025) binder DNA for Golden Gate assembly.

PyPI package: grasp-library-designer
Import name: grasp_library

Two Colab Forms notebooks:

Hard constraints (library path): protein sequence fixed (synonymous codons only); coding Golden Gate overhang bases stay locked in coding_mask. Objectives: ligation fidelity (Potapov / GGAssembler), codon optimality, synthesis fitness.

License: AGPL-3.0 (required by the vendored GGAssembler / dawdlib ligation engine). See LICENSE and THIRD_PARTY_LICENSES.md.

Install (PyPI)

pip install grasp-library-designer
# optional notebook extras
pip install "grasp-library-designer[notebook]"

Minimal API check:

from grasp_library import materialize_project, build_default_config, LigationFidelityCalculator

project = materialize_project()          # creates ./grasp_library_project + GenBank
config = build_default_config(project / "input")
print(LigationFidelityCalculator(25, 18).set_fidelity(["AATG", "GATA"]))

Until the package is published on PyPI, install from GitHub (private repo needs a PAT):

pip install "git+https://<TOKEN>@github.com/JustABiologist/grasp-library-designer.git@main"

Or clone and install editable:

git clone https://github.com/JustABiologist/grasp-library-designer.git
cd grasp-library-designer
python3 -m venv .venv
source .venv/bin/activate
pip install -e ".[notebook,dev]"

Run in Google Colab

Both notebooks use Colab Forms (#@title / #@param, {display-mode: "form"}).

1. Open a notebook

From GitHub → Open in Colab, or upload the .ipynb.

2. Install

In 0 · Install, choose:

3a. One-shot

Settings → Preview binder → Design oligos → Export Excel
Outputs: grasp_library_project/output/oneshot/{RNA}/

3b. Combinatorial library

Settings → Import → Redesign overhangs → Anneal → Pareto plot → Export → Compile target
Outputs: grasp_library_project/output/

Bundled Farley et al. GenBank modules are copied into the project folder on first run via materialize_project().

Run locally (Cursor / Jupyter / VS Code)

pip install -e ".[notebook]"
python -m ipykernel install --user --name grasp-library-designer --display-name "grasp-library-designer"

Select that kernel, open either notebook, run top-to-bottom.

Package layout

grasp_library/                 # installable Python package
  data/profiles/.../genbank/   # bundled GRASP GenBank modules
  paths.py                     # materialize_project()
  ...
third_party/dawdlib_golden_gate/   # Potapov ligation fidelity (AGPL; also installed)
grasp_*_designer.ipynb             # Colab Forms UIs (also in sdist)

Build / publish (maintainers)

pip install -e ".[dev]"
python -m build
twine check dist/*
# Test PyPI first (recommended):
twine upload --repository testpypi dist/*
# Production:
twine upload dist/*

Requires a PyPI API token (TWINE_USERNAME=__token__, TWINE_PASSWORD=pypi-...).

Hiding code (Colab / Jupyter / VS Code)

Hide ≠ protect. Source remains in the .ipynb.

License notes

• Distributed package license: AGPL-3.0 (see LICENSE).
• Vendored ligation engine under third_party/dawdlib_golden_gate/ is AGPL-3.0 (Fleishman-Lab / GGAssembler).
• GRASP sequences: Farley et al., Nucleic Acids Res. 2025.

Quick smoke test

python - <<'PY'
from pathlib import Path
from grasp_library import (
    materialize_project,
    build_default_config,
    run_oneshot_design,
    LigationFidelityCalculator,
)
from grasp_library.codon_tables import apply_organism_codon_table, load_codon_usage

project = materialize_project()
input_dir = project / "input"
cfg = build_default_config(input_dir)
cfg["optimizer"]["iterations_per_part"] = 200
apply_organism_codon_table("Escherichia coli (Kazusa)", input_dir / "codon_usage.csv")
_, codon_data = load_codon_usage(input_dir / "codon_usage.csv", genetic_code=1)
run_oneshot_design(
    target_rna="UUACACGUG",
    codon_data=codon_data,
    config=cfg,
    output_dir=project / "output" / "oneshot" / "UUACACGUG",
    n_fragments=4,
    fidelity=LigationFidelityCalculator(25, 18),
)
print("ok")
PY