Node2Vec on Hi-C: Detecting Chromosome Translocations with Network Analysis 🧬

Introduction

In the realm of genomics, unraveling the intricacies of chromosome translocations holds paramount importance, particularly in the context of understanding genetic diseases, such as cancer. This repository introduces a simple and intuitive framework that combines graph theory and machine learning to detect translocation, given an Hi-C input data.

This repository is not designed as a genomic data analysis library. Instead, it relies on the Node2Vec and scikit-learn libraries, saving you time on the data engineering part. Think of it as a tool tailored for visualizing (possible) genomic translocations when you have both a metadata file and Hi-C data at your disposal.

Requirements

Data requirements

• an Hi-C data in .csv that will be interpreted as an adjacency matrix of a graph:

  0      31495  12592  8465   5828   ...
  31495  0      55588  32895  21299  ...
  12592  55588  0      57642  32376  ...
  8465   32895  57642  0      42695  ...
  5828   21299  32376  42695  0      ...
  ...    ...    ...    ...    ...    ...
  

• a Metadata file in .csv or in .xlsx, .xls that gives informations about nodes of the graph and the related chromosome:

  chr	start	end
  chr1	1	250
  chr6	251	422
  ...	...	...

Required libraries

See the requirements.txt file.

Installation

Open a terminal or command prompt. Navigate to your preferred directory where you want to clone the repository and run:

git clone https://github.com/keivan-amini/Node2Vec-Hi-C

Move to the cloned directory and install the required libraries:

cd Node2Vec-Hi-C
python3 -m pip install -r requirements.txt .

Tutorial

Click here to access a brief tutorial with fake datasets.

Repository structure

The repository is composed by different folders:

• data contains Hi-C data and Metadata file.
• Node2VecHiC contains all python modules contained in the library.
• scripts contains the scripts used to produce the plots presented in pdf folder.
• model contains the outputs embeddings from the Node2Vec algorithm.
• plot contains the plotted figures.
• tutorial contains an easy how-to notebook exploiting some functions of the repository.
• tests contains some tests for the modules.
• pdf contains the full project documentation, with mathematical theory, figures and results.

Tests

Be careful: pytest requires: Python 3.7+ or PyPy3. If not yet installed, open a terminal and execute:

pip install -U pytest

Move to the tests directory and run the command:

cd tests
python3 -m pytest

Plot examples

References and thanks

This repository contains the project for the course of Complex Networks, part of the MSc in Applied Physics at the University of Bologna. All the references for the project are contained in the .pdf document. Special thanks to Daniel Remondini and Alessandra Merlotti!