Predicting ICB Therapy Response Using Integrative Data Sources

Project Overview

This project aims to develop a machine learning framework that predicts the response to Immune Checkpoint Blockade (ICB) therapy by integrating various genomic, transcriptomic, and clinical data sources. The model utilizes data such as gene expression, SNP mutation, and clinical data to enhance the prediction accuracy.

Features

• Integration of multiple genomic and transcriptomic data sources
• Implementation of feature selection techniques such as Recursive Feature Selection and Fischer Discriminant Ratio.
• Use of machine learning models like Random Forest, Support Vector Machines, and Logistic Regression.
• Evaluation metrics including cross-validation accuracy, testing accuracy, F1 score, and ROC AUC score.

Getting Started

Prerequisites

• Python 3 and above
• Pandas
• Numpy
• Scikit-learn
• Matplotlib
• Scipy

Installation

Clone the repository to your local machine:

• git clone https://github.com/ahmedhesham47/Machine-Learning-Integrative-Framework-for-Predicting-ICB-Response/

Usage

Navigate to the project directory and run the Jupyter notebook:

• cd path/to/your-repository
• jupyter notebook

Data

The project uses three main types of data:

• Gene Expression Data
• SNP Mutation Data
• CNA Mutation
• Clinical Data
• The datasets (Liu and Ravi) are provided in this repository in the Data Folder.

Models

The project explores several machine-learning models:

• RandomForestClassifier
• LogisticRegression
• Support Vector Machine (SVM)

Contributing

Contributions to this project are welcome. Please fork the repository and submit a pull request.