/AFnetproximity

Primary LanguagePythonMIT LicenseMIT

Transcriptomics-based Network Medicine Discovery and Population-based Validation Identifies Metformin as a Candidate Drug for Atrial Fibrillation

Network proximity code

The proximity code can be found in the proximity/ folder. First, decompress the file HumanInteractome.7z in the same folder.

  • HumanInteractome.tsv - the human interactome, including sources and evidence types
  • HumanInteractome.npy - numpy matrix of all precalculated shortest distance in the interactome
  • DrugTargetNetwork.txt - drug target network
  • network_proximity.py - code to compute closest network proximity. See below

The network_proximity.py supports two modes. To run the program (Python 3), numpy and networkx need to be installed.

pip install numpy networkx

To compute the closest proximity between two gene lists

python network_proximity.py path_to_gene_list_1 path_to_gene_list_2 number_of_repeat random_seed

To screen all the drugs for a gene list

python network_proximity.py DRUG path_to_gene_list number_of_repeat random_seed

Gene list file format

Save the gene Entrez IDs in a single column, e.g.,

58
162
166
310
373
396
409
427
476
481
514