Personalized medicine has gain much attention in recent years with the drastic reduction in genome sequencing prices and systems genetics approaches such as GWAS or PheWAS have proven to be suitable tools for genome analysis. Only recently, the use of graph methods have been gaining popularity due to their inherent flexibility and high representative power. In this project we focus on the BXD mouse genetic dataset and use network properties as well as graph signal processing method to select and infer a phenotype based on the underlying genotype and/or protein expression levels.
Find the notebook and the utils in the src folder
To run the notebook, you need the following libraries:
- numpy
- pandas
- sklearn
- pygsp
- networkx
- matplotlib
- Gianni Giusto
- Yann Metha
- Raphaël Reis Nunes
- Lucas Zweili
This project is licensed under the MIT License - see the LICENSE.md file for details