Metadata-Version: 2.4
Name: poseigen
Version: 0.1.0
Summary: Meta package for the Poseigen ML suite
Author-email: Husam Abdulnabi <husam.abdulnabi@gmail.com>
Maintainer-email: Husam Abdulnabi <husam.abdulnabi@gmail.com>
License-Expression: MIT
Project-URL: Homepage, https://github.com/husam94/poseigen
Project-URL: Repository, https://github.com/husam94/poseigen
Project-URL: Issues, https://github.com/husam94/poseigen/issues
Keywords: poseigen,bioinformatics,machine-learning,deep-learning
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Developers
Classifier: Topic :: Scientific/Engineering
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.12
Requires-Python: >=3.12
Description-Content-Type: text/markdown
Requires-Dist: poseigen_seaside
Requires-Dist: poseigen_chisel
Requires-Dist: poseigen_compass
Requires-Dist: poseigen_binmeths
Requires-Dist: poseigen_trident
Requires-Dist: poseigen_mimic
Requires-Dist: poseigen_oyster

# Poseigen

Poseigen is a modular ecosystem of Python packages for machine-learning workflows with biological sequence data. This repository is its public landing page and meta-package source; each component's source is maintained separately.

## Component packages

- [poseigen_seaside](https://github.com/Husam94/poseigen_seaside) ([PyPI](https://pypi.org/project/poseigen_seaside/)) — shared utilities, metrics, and visualization.
- [poseigen_chisel](https://github.com/Husam94/poseigen_chisel) ([PyPI](https://pypi.org/project/poseigen_chisel/)) — biological data processing and conversion into ML-ready inputs.
- [poseigen_compass](https://github.com/Husam94/poseigen_compass) ([PyPI](https://pypi.org/project/poseigen_compass/)) — hyperparameter search, candidate scoring, and ensemble evaluation.
- [poseigen_binmeths](https://github.com/Husam94/poseigen_binmeths) ([PyPI](https://pypi.org/project/poseigen_binmeths/)) — binning, balanced sampling, and split-generation methods.
- [poseigen_trident](https://github.com/Husam94/poseigen_trident) ([PyPI](https://pypi.org/project/poseigen_trident/)) — neural-network building blocks and training utilities.
- [poseigen_mimic](https://github.com/Husam94/poseigen_mimic) ([PyPI](https://pypi.org/project/poseigen_mimic/)) — unified sequence-embedding architectures built on Trident.
- [poseigen_oyster](https://github.com/Husam94/poseigen_oyster) ([PyPI](https://pypi.org/project/poseigen_oyster/)) — glass-box genomic sequence models with exact position-specific k-mer contributions.

## How the packages fit together

- Seaside provides shared utilities.
- Chisel prepares biological data for modeling.
- Binmeths provides binning and data-splitting methods.
- Trident provides reusable neural-network and training components.
- Mimic and Oyster build sequence models using Trident.
- Compass supports model selection, optimization, and evaluation.

## Installation

Install the full suite from PyPI:

```bash
python -m pip install poseigen
```

You can also install individual components, for example:

```bash
python -m pip install poseigen_binmeths poseigen_oyster poseigen_trident
```

See each component's repository for its documentation, source, and release history.

## Development and releases

This is a hub, not a monorepo; component source code is not vendored here. See [RELEASE.md](RELEASE.md) for the suite's release process.
