Metadata-Version: 2.4
Name: poseigen_oyster
Version: 1.0.0
Summary: The basics for Poseigen software
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_oyster
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Developers
Classifier: Topic :: Software Development
Classifier: Programming Language :: Python :: 3
Requires-Python: >=3.12
Description-Content-Type: text/markdown
License-File: LICENSE.txt
Requires-Dist: numpy
Requires-Dist: torch
Requires-Dist: poseigen_seaside
Requires-Dist: poseigen_trident
Requires-Dist: poseigen_chisel
Requires-Dist: scikit-learn
Requires-Dist: kneed
Dynamic: license-file

# Oyster

Oyster is a Neural Network architecture that can be toggled to consider (non-Exact) or ignore (Exact) k-mer associations. 

This package implements Oyster. It has two modules: 

1. model
- Here you will find the Oyster model. It is built using the [Trident](https://github.com/Husam94/poseigen_trident).
- Dual Oyster combines two Oyster models to model different types of biological sequences. It is exemplified by modelling [YY1 interactions](https://github.com/Husam94/YY1) using genomic DNA sequences and histone signals. Dual Oyster also serves as a template for combining more than one Oyster into a single model. 
- Recommended search spaces for hyperparameter optimizations are provided in [poseigen_compass](https://github.com/Husam94/poseigen_compass) format
- Reset functions are provided to initialize or reinitialize models for model fitting.

2. analysis 
- Here you will find functions to analyze Exact Oyster architectures, to uncover learned concepts. These functions include determining the contribution of each sequence (for multi-sequence inputs), k-mer contributions, k-ppm contributions, and k-mer clustering and refining. 
- Follow the example in modelling [YY1 interactions](https://github.com/Husam94/YY1). 
