Introduction

This repository hosts Hao Chen's coursework for LPI Data Science module 2023-2024. The project entails an analysis of disease-to-disease network and disease-to-gene network to compare syndromic and non-syndromic autism. It used gene-disease association data collected from the DisGeNet database.

The analysis result is summarised in a final slide presentation, stored in the presentation folder. The data analysis python notebook is stored in the notebook folder.

Data Analysis Logic

For a quick introduction to each notebook, please see below:

• 0_autism_terms.ipynb: select autism disorder diagnosis id, result stored in notebook/processed_data/autism_terms.csv
• 1_group_autism_subtypes.ipynb: reprocess raw data by regrouping autism diagnosis according to subtypes, processed data stored in notebook/processed_data/curated_gene_disease_associations_autism_grouped.csv
• 2_compare_gene_properties.ipynb: compare gene properties for autism subtypes
• 3_create_nodes_edges_grouped.Rmd: hyper-geometric test for the p-value of disease-to-disease association, and create nodes and edges for network analysis:
  • Edge data: notebook/processed_data/edges_shared_genes_AutismGrouped.csv
  • Node data: notebook/processed_data/nodes_diseases_AutismGrouped.csv
• 4_create_network.ipynb: create disease-to-disease network:
  • Network for all diseases, stored in notebook/graphml/p3_network_all_diseases.graphml
  • Network for autism and neighbour diseases, stored in notebook/graphml/p3_network_autism_neighbors.graphml
• 5_centrality_analysis_original.ipynb: analysis of centrality measurement of disease nodes
• 6_bipartie_autism.ipynb: create disease-to-gene network for autism genes and autism subtypes
• 7_bipartie_alldisease.ipynb: create disease-to-gene network for all diseases and related genes
• 8_autism_disease_class.ipynb: analysis of disease-to-class network

Data visualization

All data visualization figures and output are stored in figures folder:

• The figures/gene_properties folder contains figures for gene comparison, the code for generating these figures is contained in notebook/2_compare_gene_properties.ipynb
• The figures/network folder contains figures for network visualisation, which was conducted using Gephi. The Gephi files and network data is stored in the notebook/graphml folder.