Produce JSON files from GenomeChronicler output

produce_react_files.py

import os
import csv
import json
import re
import sqlite3
import sys
from shutil import copyfile
import html

def execute_db_query(query, data):
  with DB_CONNECTION:
    cur = DB_CONNECTION.cursor()
    cur.execute(query, data)
    rows = cur.fetchall()
    if len(rows) == 1:
      return html.unescape(rows[0][0].strip().capitalize())
    else:
      return ''

def get_trait_full_description(id, allele_1, allele_2):
  query = "SELECT summary FROM data where id = ? AND allele1 = ? and allele2 = ? "
  execute_db_query(query, (id, allele_1, allele_2))
  
def get_genoset_full_description(id):
  query = "SELECT summary FROM genoset where id = ?"
  execute_db_query(query, (id,))

def process_link(text):
  matches = re.findall(r'{(.+)}{(.+)}', text)
  if matches == []:
    links = ['', '']
  else:
    links = list(matches[0])
  return links

def convert_genoset_tables_to_json(input_file, output_file):
  json_data = []
  with open(input_file) as csvfile:
    next(csvfile)
    data = csv.reader(csvfile)
    for row in data:
      identifier_obj = process_link(row[1])
      description_obj = process_link(row[2])
      row_data = {
        "magnitude": row[0],
        "identifer": {
          "link": identifier_obj[0],
          "name": identifier_obj[1]
        },
        "description": {
          "short": description_obj[1],
          "full": get_genoset_full_description(identifier_obj[1])
        },
        "db_links": [
          { "name": "SNPedia", "link": identifier_obj[0] },
        ]
      }
      json_data.append(row_data)
  with open(output_file, 'w') as f:
    json.dump(json_data, f, sort_keys=True, indent=4)

def convert_trait_tables_to_json(input_file, output_file):
  json_data = []
  with open(input_file) as csvfile:
    next(csvfile)
    data = csv.reader(csvfile)
    for row in data:
      identifier_obj = process_link(row[1])
      description_obj = process_link(row[3])
      gnomad_obj = process_link(row[4])
      get_evidence_obj = process_link(row[5])
      clinvar_obj = process_link(row[6])
      alleles = list(re.findall(r'([-\w]);([-\w])', row[2])[0])
      row_data = {
        "magnitude": row[0],
        "identifer": {
          "link": identifier_obj[0],
          "name": identifier_obj[1]
        },
        "genotype": alleles,
        "description": {
          "short": description_obj[1],
          "full": get_trait_full_description(identifier_obj[1], alleles[0], alleles[1])
        },
        "db_links": [
          { "name": "SNPedia", "link": identifier_obj[0] },
          { "name": "GnomAD", "link": gnomad_obj[0]},
          { "name": "GetEvidence", "link": get_evidence_obj[0]},
          { "name": "ClinVar", "link": clinvar_obj[0]}
        ]
      }
      json_data.append(row_data)
  with open(output_file, 'w') as f:
    json.dump(json_data, f, sort_keys=True, indent=4)


db = r"/data/pgp_data/snpedia.db"
DB_CONNECTION = sqlite3.connect(db)
DB_CONNECTION.text_factory = lambda x: str(x, 'latin1')

data_dir = sys.argv[1]

output_folder = os.path.join(data_dir, "react_data")
os.mkdir(output_folder)

copyfile(os.path.join(data_dir, 'AncestryPlot.pdf'),
         os.path.join(output_folder, 'AncestryPlot.pdf'))
convert_trait_tables_to_json(
  os.path.join(data_dir, 'latest.good.reportTable.csv'),
  os.path.join(output_folder, 'beneficial_data.json'))
convert_trait_tables_to_json(
  os.path.join(data_dir, 'latest.bad.reportTable.csv'),
  os.path.join(output_folder,'harmful_data.json'))
convert_genoset_tables_to_json(
  os.path.join(data_dir, 'latest.genoset.reportTable.csv'),
  os.path.join(output_folder,'genosets_data.json'))