EEI-Networks: Conservation-Based Exon-Exon Interaction Detection

This repository contains a comprehensive four-pipeline approach for constructing and expanding Human Exon-Exon Interaction (EEI) networks through multiple complementary methods. The work extends the original research on cancer-related protein complex interface aberrations by leveraging structural data from multiple eukaryote species.

Project Overview

This thesis project implements a comprehensive approach to expand human EEI networks beyond direct detection by leveraging evolutionary conservation patterns across multiple species. The project includes:

Pipeline Structure

1. Human Baseline (EEI-Homo-Sapiens/)

• EEI-networks of Homo Sapiens built using UniProt data updated to October 2024
• Three complementary detection methods: Contact-based, PISA-based, and EPPIC-based
• High-confidence combined network generation

2. Species Networks (EEI-[Species]/)

• EEI networks for 7 eukaryote species: Mouse, Cattle, Fruit fly, Chicken, Rabbit, Rat, and Yeast
• Each species includes EGIO orthology detection
• Species-specific orthology-based EEI prediction analysis in human_EEI_prediction/ folders
• High-confidence network generation for each species

3. Species-Specific Orthology Analysis (EEI-[Species]/human_EEI_prediction/)

• Orthology-based EEI prediction algorithms for each species
• Analysis and evaluation tools for prediction accuracy
• Results from multiple detection methods (Contact, PISA, EPPIC)

4. Comprehensive Analysis (final_statistics/)

• EEI network statistics and method comparisons
• EGIO orthology analysis and evolutionary divergence studies
• Phylogenetic analysis and network visualizations
• Overlap analysis and Venn diagram generation

5. Mouse RNA-seq Analysis (mm_RNA_seq_data/)

• Survival analysis based on EEI networks
• Expression correlation analysis
• Cancer-related protein expression studies
• Treatment response analysis

Usage

Each folder contains numbered scripts that should be executed in ascending order. For detailed instructions, refer to:

• THESIS_DOCUMENTATION.md - Comprehensive project documentation
• PROJECT_SUMMARY.md - Executive summary
• TECHNICAL_IMPLEMENTATION.md - Technical implementation details

Key Features

• Multi-Method Integration: Combines three complementary EEI detection approaches
• Evolutionary Conservation: Leverages orthology to expand human EEI networks
• Cross-Species Validation: Uses multiple species for robust predictions
• Comprehensive Analysis: Statistical evaluation and visualization of results
• Web Interface: EEINet database for network exploration

Web Interface

The EEINet web interface provides an interactive platform for exploring and analyzing the EEI networks generated by this project:

• Repository: EEINet GitHub Repository
• Live Application: EEINet Web Interface

The web interface allows users to:

• Browse and search EEI networks across different species
• Visualize network interactions and conservation patterns
• Access detailed statistics and analysis results
• Download network data and analysis outputs