Bayesian Neural Network Biological Network Inference

This BNN BNI Analysis Tool can be used for predictive insight in Biological Networks through the use of Bayesian Neural Networks. This repository can be used for any biological datasets, however it is tailored toward individuals who are interested in neural networks/neural computation, computational biology, computational systems biology, computational genomics, computational pharmacology, and computational neuroscience. It is also made just as available to anyone who would like to use it for their own learning endeavours or to start their journey in any of these fields!

Directory Structure

BNN_Biological_Network_Inference/
├── data/
│   └── example_data.csv
│
├── src/
│   ├── data_preprocessing.py
│   ├── train.py
│   ├── inference.py
│   └── utils.py
│
├── .gitignore
├── LICENSE
├── requirements.txt
└── README.md

Example Databases

• Protein Data

• Genomic Data

• Drug Interaction Data

• Metabolite Data

• Microbiome Data

Data Structure

Your .csv files should follow this structure:

1. Features: Columns representing various features.
2. Target Variable: A column for the target variable that represents the interaction score or outcome you want to predict.

Example Structure:

feature1,feature2,feature3,target

value1,value2,value3,target_value

Instructions

1. Prepare Your Data

• Copy The Repository:

git clone https://github.com/peterajhgraham/BNN_Biological_Network_Inference.git
cd BNN_Biological_Network_Inference

• Format Your Data: Ensure your .csv files follow the structure outlined above.

• Place Data Files: Save your .csv files in the data/ directory of the repository.

2. Install Required Packages

Install the required Python packages listed in requirements.txt:

pip install -r requirements.txt

3. Preprocess Data

Run the data preprocessing script to load and prepare your data for modeling:

from src.data_preprocessing import load_data, preprocess_data

file_paths = {
    'example': 'data/example_data.csv'
}

data = load_data(file_paths)
processed_data = preprocess_data(data)

4. Train the Model

Use the training script to build and train your Bayesian Neural Network model with torchbnn:

from src.train import train_model

input_dim = processed_data['example'][0].shape[1]  # Example for one dataset
model = train_model(*processed_data['example'], input_dim)

5. Make Predictions

After training, use the inference script to make predictions with your model:

from src.inference import infer

predictions = infer(model, processed_data['example'][1])

6. Visualize Results

You can use the utility functions to visualize your results:

from src.utils import plot_predictions

plot_predictions(processed_data['example'][2], predictions)