BioNeuralNet - Multi-Omics Integration with Graph Neural Networks

Installation

To install BioNeuralNet, simply run:

pip install bioneuralnet

For additional installation details, see Installation.

BioNeuralNet Overview

Embeddings form the core of BioNeuralNet, enabling a number of downstream applications.

BioNeuralNet Core Features

For an End-to-End example example of BioNeuralNet, see Quick Start Guide.

Network Embedding: GNN Embeddings

• Given a multi-omics network as input, BioNeuralNet can generate embeddings using Graph Neural Networks (GNNs).

• Generate embeddings using methods such as GCN, GAT, GraphSAGE, and GIN.

• Outputs can be obtained as native tensors or converted to pandas DataFrames for easy analysis and visualization.

• Embeddings unlock numerous downstream applications, including disease prediction, enhanced subject representation, clustering, and more.

Graph Clustering: Correlated Clustering

• Identify functional modules or communities using correlated clustering methods (e.g., CorrelatedPageRank, CorrelatedLouvain, HybridLouvain) that integrate phenotype correlation to extract biologically relevant modules [1].

• Clustering methods can be applied to any network represented allowing flexible analysis across different domains.

• All clustering components return either raw partitions dictionaries or induced subnetwork adjacency matrices (as DataFrames) for visualization.

• Use cases include, feature selection, biomarker discovery, and network-based analysis.

Downstream Tasks: Downstream Tasks

• Subject Representation:

  • Integrate node embeddings back into omics data to enrich subject-level profiles by weighting features with learned embedding.

  • This embedding-enriched data can be used for downstream tasks such as disease prediction or biomarker discovery.

  • The result can be returned as a DataFrame or a PyTorch tensor, fitting naturally into downstream analyses.

• Disease Prediction for Multi-Omics Network DPMON [2]:

  • Classification End-to-End pipeline for disease prediction using Graph Neural Network embeddings.

  • DPMON supports hyperparameter tuning-when enabled, it finds the best for the given data.

  • This approach, along with the native pandas integration across modules, ensures that BioNeuralNet can be easily incorporated into your analysis workflows.

Metrics: Metrics

• Several plotting funcctions to visualize networks, emebddings, variance distribution, cluster comparison, and more.

• Correlation based functions to compare clustersand omics data with the phenotype.

Utilities: Utils

• Filtering Functions:

  • Network filtering allows users to select variance or zero-fraction filtering to an omics network.

  • Reducing noise, and removing outliers.

• Data Conversion:

  • Convert RData files both CSV and to Pandas DataFrame. For ease of integration for R-based workflows.

External Tools: External Tools

• Graph Construction:

  • BioNeuralNet provides additional tools in the bioneuralnet.external_tools module.

  • Allowing users to generate networks using R-based tools like SmCCNet.

  • While optional, these tools enhance BioNeuralNet’s capabilities and are recommended for comprehensive analysis.

What is BioNeuralNet?

BioNeuralNet is a Python-based framework designed to bridge the gap between multi-omics data analysis and Graph Neural Networks (GNNs). By leveraging advanced techniques, it enables:

• Graph Clustering: Identifies biologically meaningful communities within omics networks.

• GNN Embeddings: Learns network-based feature representations from biological graphs, capturing both biological structure and feature correlations for enhanced analysis.

• Subject Representation: Generates high-quality embeddings for individuals based on multi-omics profiles.

• Disease Prediction: Builds predictive models using integrated multi-layer biological networks.

Why GNNs?

Traditional methods often struggle to model complex multi-omics relationships due to their inability to capture biological interactions and dependencies. BioNeuralNet addresses this challenge by utilizing GNN-powered embeddings, incorporating models such as:

• Graph Convolutional Networks (GCN): Aggregates features from neighboring nodes to capture local structure.

• Graph Attention Networks (GAT): Applies attention mechanisms to prioritize important interactions between biomolecules.

• GraphSAGE: Enables inductive learning, making it applicable to unseen omics data.

• Graph Isomorphism Networks (GIN): Improves expressiveness in graph-based learning tasks.

By integrating omics features within a network-aware framework, BioNeuralNet preserves biological interactions, leading to more accurate and interpretable predictions. For a deeper dive into how BioNeuralNet applies GNN embeddings, see GNN Embeddings.

Seamless Data Integration

One of BioNeuralNet’s core strengths is interoperability:

• Outputs are structured as pandas DataFrames, ensuring easy downstream analysis.

• Supports integration with external tools and machine learning frameworks, making it adaptable to various research workflows.

• Works seamlessly with network-based and graph-learning pipelines.

• Our User API provides detailed information on how to use BioNeuralNet’s modules and functions.

Example: Transforming Multi-Omics for Enhanced Disease Prediction

View full-size image: Transforming Multi-Omics for Enhanced Disease Prediction

BioNeuralNet: Transforming Multi-Omics for Enhanced Disease Prediction

Below is a quick example demonstrating the following steps:

1. Data Preparation:

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

2. Network Construction:

   • Not performed internally: Generate the network adjacency matrix externally (SmCCNet).

   • Lightweight wrappers (SmCCNet) are available in bioneuralnet.external_tools for convenience, R is required for their usage.

3. Disease Prediction:

   • Use DPMON to predict disease phenotypes by integrating the network information with omics data.

   • DPMON supports an end-to-end pipeline with hyperparameter tuning that can return predictions as pandas DataFrames, enabling seamless integration with existing workflows.

Code Example:

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

# Step 1: Data Preparation
phenotype_data = pd.read_csv('phenotype_data.csv')
omics_proteins = pd.read_csv('omics_proteins.csv')
omics_metabolites = pd.read_csv('omics_metabolites.csv')
clinical_dt = pd.read_csv('clinical_data.csv')

# Step 2: Network Construction
smccnet = SmCCNet(
    phenotype_df=phenotype_data,
    omics_dfs=[omics_proteins, omics_metabolites],
    data_types=["protein", "metabolite"],
    kfold=5,
    summarization="PCA",
)
global_network, clusters = smccnet.run()
print("Adjacency matrix generated.")

# Step 3: Disease Prediction (DPMON)
dpmon = DPMON(
    adjacency_matrix=global_network,
    omics_list=[omics_proteins, omics_metabolites],
    phenotype_data=phenotype_data,
    clinical_data=clinical_dt,
    model="GCN",
)
predictions = dpmon.run()
print("Disease phenotype predictions:\n", predictions)

Contents:

• Installation
• GNN Embeddings
  • Key Contributions:
  • GNN Model Overviews
  • Dimensionality Reduction and Downstream Integration
  • Task-Driven (Supervised/Semi-Supervised) GNNs
  • Generating Low-Dimensional Embeddings for Multi-Omics
  • Key Insights into GNN Parameters and Outputs
  • Dimensionality Reduction: PCA vs. Autoencoders
  • How DPMON Uses GNNs Differently
  • Example Usage
• Correlated Clustering
  • Overview
  • Mathematical Approach
  • Graphical Comparison
  • Integration with BioNeuralNet
  • Notes for Users
  • Conclusion
• Metrics
  • Correlation Metrics
  • Visualization
  • Evaluation
  • Example Usage
  • Further Information
• Utils
  • RData Conversion
  • Logging
  • Graph Generation
  • Preprocessing Utilities
  • Data Summary Utilities
  • References
• Downstream Tasks
  • Overview
  • DPMON: Disease Prediction Pipeline
  • Subject Representation & Embedding Integration
  • Unlocking Downstream Applications
  • Get Started
  • References
• Quick Start Guide
  • BioNeuralNet offers a number of classes and functions.
  • Load Data
  • Generate Global Network using SmCCNet
  • Generate Network Embeddings using GNNEmbedding
  • Embeddings visualization
  • Integrate Embeddings into Omics Data with SubjectRepresentation
  • Comparing results for prediction task
  • Disease Prediction using DPMON(Disease Prediction using Multi-Omics Networks)
  • Phenotype after:
  • Visualizing the variance distribution and feature variance
  • DPMON Predictions
  • Clustering with CorrelatedLouvain and HybridLouvain
  • Lets plot the clustered network from correlated louvain
• TCGA-BRCA Demo
  • Dataset Source
  • Dataset Overview
  • Patients in Every Dataset
  • Final Shapes (Per-Patient)
  • Data Summary Table
  • Reference Links
  • TCGAbiolinks
• Preprocessing
• Optional: Load the data we just saved to make sure it looks okay.
• Easy Access via DatasetLoader
• Preparing Multi-Omics Data for downstream tasks
• TOPMed Demo
  • Trans-Omics for Precision Medicine | TOPMed
  • Loading your data:
  • Ease of component exploration via DatasetLoader
  • Generating a Multi-Omics Network using SmCCNet
  • SmCCNet Output
  • Generating Low-Dimensional Embeddings using Graph Neural Networks to capture meaningful biological interactions.
  • Output
  • Embeddings visualization
  • Using the Embeddings
  • Comparing results
  • Network Visulization
  • Correlated Clustering
  • DPMON (Disease Prediction using Multi-Omics Networks) reuses the same GNN architectures but with a different objective:
  • DPMON allows BioNeuralNet users to significatly improve phenotype predictions with a few lines of code.
• Tutorials
  • Example 1: SmCCNet + GNN Embeddings + Subject Representation
  • Example 2: SmCCNet + DPMON for Disease Prediction
  • Graph Embeddings
  • Subject Representation
  • Disease Prediction (DPMON)
  • Disease Prediction
  • Graph Clustering
• External Tools
  • Graph Construction
• User API
  • bioneuralnet.network_embedding
  • bioneuralnet.clustering
  • bioneuralnet.downstream_task
  • bioneuralnet.utils
  • bioneuralnet.datasets
  • bioneuralnet.metrics
  • bioneuralnet.external_tools
  • Executables
• Acknowledgments
  • Key Dependencies
  • Contributors
  • Frequently Asked Questions (FAQ)

Indices and References

• Index

• Module Index

• Search Page

[1]

Abdel-Hafiz, M., Najafi, M., et al. “Significant Subgraph Detection in Multi-omics Networks for Disease Pathway Identification.” Frontiers in Big Data, 5 (2022). DOI: 10.3389/fdata.2022.894632.

[2]

Hussein, S., Ramos, V., et al. “Learning from Multi-Omics Networks to Enhance Disease Prediction: An Optimized Network Embedding and Fusion Approach.” In 2024 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), Lisbon, Portugal, 2024, pp. 4371-4378. DOI: 10.1109/BIBM62325.2024.10822233.