.. BioNeuralNet documentation master file

Welcome to BioNeuralNet Beta 0.1
================================

**Note:** This is a **beta version** of BioNeuralNet. It is under active development, and certain features 
may be incomplete or subject to change. Feedback and bug reports are highly encouraged to help us 
improve the tool.

BioNeuralNet is a Python-based software tool designed to streamline the transformation of multi-omics 
data into network-based representations and lower-dimensional embeddings, enabling advanced analytical 
processes like clustering, feature selection, disease prediction, and environmental exposure assessment.
**Python Installation via pip:**

   .. code-block:: bash

      pip install bioneuralnet==0.1.0b1
      
For installation details and other options, go to :doc:`installation`.

**Example: Transforming Multi-Omics for Enhanced Disease Prediction** 
---------------------------------------------------------------------

`View full-size image: Transforming Multi-Omics for Enhanced Disease Prediction <https://ramosv.github.io/_images/Overview.png>`_

.. figure:: _static/Overview.png
   :align: center
   :alt: BioNeuralNet Overview

   **BioNeuralNet**: Transforming Multi-Omics for Enhanced Disease Prediction


BioNeuralNet enables seamless integration of multi-omics data into a network-based analysis pipeline. 
Here is a quick example demonstrating how to generate a network representation using SmCCNet and apply it 
to disease prediction using DPMON:

**Note**: 

1. **Data Preparation**:

   - Input your multi-omics data (e.g., proteomics, metabolomics, genomics) along with phenotype data.

2. **Network Construction**:

   - Use Sparse Multiple Canonical Correlation Network (SmCCNet) to generate a network from the omics data. 
   - This step constructs an adjacency matrix capturing correlations and interactions between features.

3. **Disease Prediction**:

   - Disease Prediction using Multi-Omics Networks (DPMON) uses Graph Neural Networks (GNNs) to predict disease phenotypes.
   - Integrates multi-omics data and network structure information to generate GNNs embeddings that capture global and local graph patterns.
   - It enhances the Omics-data by creating enriched with node features. These are processed through a Neural Network, optimized end-to-end, enhancing predictive accuracy and reducing overfitting.

**Code Example**:

.. code-block:: python

   import pandas as pd
   from bioneuralnet.graph_generation import SmCCNet
   from bioneuralnet.downstream_task import DPMON

   # Step 1: Load Multi-Omics Dataset
   omics_data = pd.read_csv('omics_data.csv', index_col=0)
   phenotype_data = pd.read_csv('phenotype_data.csv', index_col=0)

   # Step 2: Generate a network using SmCCNet
   smccnet = SmCCNet(phenotype_data=phenotype_data, omics_data=omics_data)
   adjacency_matrix = smccnet.run()
   print("Multi-Omics Network generated.")

   # Step 3: Enhanced disease prediction using network information with DPMON
   dpmon = DPMON(adjacency_matrix=adjacency_matrix, omics_list=[omics_data], phenotype_data=phenotype_data)
   predictions = dpmon.run()
   print("Disease phenotype predictions:")
   print(predictions)

**Output**:

  - **Adjacency Matrix**: The network representation of the multi-omics data.
  - **Predictions**: Disease phenotype predictions for each sample.


**BioNeuralNet Overview**
--------------------------
Looking to explore the capabilities of BioNeuralNet? Here is a brief overview of the key components:

1. **Graph Construction**: Build multi-omics networks using methods like Weighted Gene Co-expression Network Analysis (**WGCNA**), Sparse Multiple Canonical Correlation Network (**SmCCNet**), or import existing networks.
2. **Graph Clustering**: Identify functional modules and communities using hierarchical clustering, PageRank, or Louvain clustering algorithms.
3. **Network Embedding**: Generate embeddings with **Graph Neural Networks (GNNs)** or **Node2Vec**, simplifying high-dimensional data into lower-dimensional representations.
4. **Subject Representation**: Integrate embeddings into omics data to enrich subject-level features, enhancing the dataset for downstream analyses.
5. **Downstream Tasks**: Perform advanced analyses like disease prediction using network information. Seamlessly integrate your own downstream tasks by leveraging existing components.

`View full-size image: BioNeuralNet Overview <https://ramosv.github.io/_images/BioNeuralNet.png>`_

.. figure:: _static/BioNeuralNet.png
   :align: center
   :alt: BioNeuralNet

   BioNeuralNet Overview

**Subject Representation**

`View full-size image: Subject Representation <https://ramosv.github.io/_images/SubjectRepresentation.png>`_

.. figure:: _static/SubjectRepresentation.png
   :align: center
   :alt: Subject Representation Workflow

   Subject-level embeddings provide richer phenotypic and clinical context.

**Disease Prediction**

`View full-size image: Disease Prediction (DPMON) <https://ramosv.github.io/_images/DPMON.png>`_

.. figure:: _static/DPMON.png
   :align: center
   :alt: Disease Prediction (DPMON)

   Embedding-enhanced subject data using DPMON for improved disease prediction.

Documentation Overview
-----------------------

.. toctree::
   :maxdepth: 2
   :caption: Contents:

   installation
   tutorials/index
   tools/index
   user_api
   faq


Indices and tables
==================

* :ref:`genindex`
* :ref:`modindex`
* :ref:`search`