Filter genes for whitelist on http://tintori.bio.unc.edu

filter_genes.py

#/usr/bin/env python

import os
import json 
import sys
import subprocess

whitelist_url = 'https://raw.githubusercontent.com/ptgolden/ma-browser/master/src/gene_whitelist.json'

if __name__ == '__main__':
    if not os.path.exists('gene_whitelist.json'):
        subprocess.call(['wget', '-nc', whitelist_url]) 

    with open('gene_whitelist.json', 'r') as fp:
        whitelist = { gene for gene in json.load(fp) }

    try:
        csv_file = sys.argv[1]
    except:
        raise EnvironmentError('First argument must be a list of genes to filter')

    genes = [
        gene for gene in open(csv_file, 'r')
        if gene.split(',')[0] in whitelist
    ]   

    print(''.join(genes).strip())