Python script to query the AskGalaxy service to generate CSV data for Observable.

ask-galaxy.py

#!/usr/bin/env python3

import os
import sys
import json
import requests

KB = 1024
MB = KB * KB
GB = MB * KB

service = 'https://ask.galaxyproject.org'
tools = ['bowtie2', 'bwa_mem', 'minimap2' ]
metric = sys.argv[1]

def get_url(tool, metric, input_size):
	size = input_size * GB
	params = ''
	if tool == 'bowtie2':
		params = f'fastqsanger_file_size_bytes_1={size}&fastqsanger_file_size_bytes_2={size}&fasta_file_size_bytes=1000000&format_output=false'	
	elif tool == 'bwa_mem' or tool == 'minimap2':
		p = []
		for x in [1,2,11,12]:
			p.append(f"fastq_input{x}={size}")
		params = '&'.join(p) + "&ref_file=1000000"
	elif tool == 'featurecounts':
		params = f'alignment_file_size_bytes={size}'
	else:
		raise Exception(f'Unsupported tool type {tool}')
	return f'{service}/{tool}/{metric}/?{params}'
	
		
headers = { 'Accept': 'application/json' }
print(f"GB,Tool,Value")

for i in range(30):
	size = i+1
	for tool in tools:
		url = get_url(tool, metric, size)
		#print(url)
		response = requests.get(url, headers=headers)
		data = response.json()
		if 'memory' == metric:
			print(f"{size},{tool},{(data['Required memory']/GB):02.2f}")
		else:
			print(f"{size},{tool},{(data['Required CPU']):02.2f}")