Download results from One Codex and combine into a single table (python) (note: requires pandas)

download_result_table.py

#!/usr/bin/python
"""Download results from One Codex and combine into a single table (note: requires pandas)."""

import requests
import os
import pandas as pd

# Expects the API key for One Codex to be stored in the ONE_CODEX_API_KEY environment variable
api_key = os.environ['ONE_CODEX_API_KEY']

# Set authentication for requests
auth = (api_key, '')

# Get the complete set of analyses
analyses = requests.get('https://app.onecodex.com/api/v0/analyses', auth=auth)
analyses = analyses.json()

dat = {}
for a in analyses:
    # Only keep completed analyses with the One Codex Database
    if a['reference_name'] == 'One Codex Database' and a['analysis_status'] == 'Success':
        url = 'https://app.onecodex.com/api/v0/analyses/%s/extended_table' % (a['id'])
        results = requests.get(url, auth=auth)
        results = results.json()
        # Convert to a Pandas DataFrame
        results = pd.DataFrame(results)
        # Index data by name, rank, and tax_id
        results.set_index(['name', 'rank', 'tax_id'], inplace=True)
        # Store the `readcount_w_children` variable, keyed by sample_filename and sample_id
        dat[(a['sample_filename'], a['sample_id'])] = results['readcount_w_children']

# Format the complete result table (replacing NaN's with 0 values)
dat = pd.DataFrame(dat).fillna(0)

# Write out the result table to a file
dat.to_csv('results_table.csv')