bsseq-run

#!/usr/bin/env python3
# -*- coding: utf-8 -*-

import os
import subprocess
import tempfile
import logging
import sys
import re


VERSION = "0.1"


if sys.version_info[0] < 3:
    print("This script requires Python 3+")
    exit(1)

logging.basicConfig(level=logging.INFO, 
                    format='%(asctime)s : %(levelname)s : %(message)s',
                    datefmt='%H:%M:%S')
logger = logging.getLogger(__name__)


import argparse
import json

def is_in_path(path):
    try:
        if subprocess.run(["which",path],stdout=subprocess.PIPE,stderr=subprocess.STDOUT).returncode != 0:
            return False
    except:
        return False
    return True

def run_shell(cmd):
    status = 0
    try:
        output = subprocess.check_output(cmd, shell=True, stderr=subprocess.STDOUT)
    except subprocess.CalledProcessError as e:
        status = e.returncode
        output = e.output
    output = output.decode()
    return (status, output)

if not is_in_path("samtools") and not is_in_path("fastp") and not is_in_path("cgmaptools") and not is_in_path("java"):
    exit("please check the env if have samtools cgmaptools fastp")
    
def main():
    parser = argparse.ArgumentParser(
        description="BS-seq analysis workflow base on bsseeker and CGmaptools"
    )

    parser.add_argument("--input","-i",type=str, help="the input json", required=True)

    parser.add_argument(
        "--cromwell","-c", help="Optional path to cromwell jar file",
    )

    parser.add_argument("--version", "-v", action='store_true', help="show version tag: {}".format(VERSION))

    args = parser.parse_args()

    if args.version:
        exit(VERSION)

    cromwell = args.cromwell
    input_json = args.input
    base_dir = os.path.dirname(os.path.realpath(__file__))
    wdl_dir = os.path.join(base_dir, "WDL")
    meta_file = "bsseq.output.json"
    
    cromwell_jar = None
    if cromwell is None:
        for f in os.listdir(wdl_dir):
            if f.startswith('cromwell-') and f.endswith('.jar'):
                cromwell_jar = os.path.join(wdl_dir, f)
                break
    else:
        cromwell_jar = os.path.abspath(cromwell)
    if cromwell_jar is None or not os.path.exists(cromwell_jar):
        exit(
            "Cromwell jar file not found. Please run download_cromwell.sh to download."
        )

    if input_json is None and not os.path.exists("bsseq.input.json"):
        exit("Need the input parameter with json files")
    else:
        input_json = "bsseq.input.json"

    cromwell_args = [
        "java",
        "-jar",
        cromwell_jar,
        "run",
        "-i",
        input_json,
        '-m',
        meta_file,
        os.path.join(wdl_dir, "bsseq.wdl"),
    ]

    logger.info("CMD: {}".format(" ".join(cromwell_args)))
        
    subprocess.run(cromwell_args)

if __name__ == "__main__":
    if "--version" in sys.argv:
        print(f"ctat-vif v{VERSION}")
        sys.exit(0)
    
    main()