asodesigner 0.1.4

Feature extraction and analysis toolkit for antisense oligonucleotide design.

ASOdesigner

ASOdesigner provides feature extraction utilities that support the design of antisense oligonucleotides (ASOs). The package bundles folding, accessibility, and sequence analysis helpers that were originally created for the TAU-Israel 2025 iGEM project.

Installation

You can install the library directly from PyPI once it is published:

pip install asodesigner

For local development and experimentation inside this repository, install the package in editable mode together with its runtime dependencies:

pip install -r requirements.txt

Usage

Functions are organised by module. For example, to compute a few sequence-derived features:

from asodesigner.features import seq_features

gc = seq_features.get_gc_content("ATGGCC")

Refer to the individual module docstrings for detailed behaviour and expected inputs.

Development Workflow

1. Update or add functionality under src/asodesigner/.
2. Keep imports relative within the package (for example, use from .util import helper).
3. Run python -m compileall src/asodesigner or your preferred test suite to make sure the code still imports correctly.

Genome Cache

asodesigner.genome.get_hg38_genome_path() downloads hg38 into cache/genomes/hg38 when the file is first requested and returns the local FASTA path. Set ASODESIGNER_HG38_PATH if you already have a FASTA on disk, or ASODESIGNER_HG38_URL to use a different mirror.

Docker Compose Workflow

docker-compose.yml keeps the genome in a named volume so large files stay out of the repository. Start the stack with the default curl download:

docker compose up

Switch to the simple Python downloader by setting GENOME_FETCH_METHOD=python. Both services mount the genome-cache volume, so subsequent runs reuse the same FASTA file.

Building and Publishing to PyPI

The project uses a modern pyproject.toml driven build. To prepare a release:

1. Choose a version – bump the version field in pyproject.toml following semantic versioning.
2. Build the artifacts:

   python -m build
   

3. Inspect the build – verify the generated wheel and source distribution under dist/.
4. Upload to PyPI (or TestPyPI first) using twine:

   python -m twine upload dist/*
   

Remember that publishing requires a PyPI account and API token. For TestPyPI, swap the repository URL accordingly.

License

The code is released under the Creative Commons Attribution 4.0 International License (CC BY 4.0). See LICENSE for full details.