run_clairs

#!/usr/bin/env python

# BSD 3-Clause License
#
# Copyright 2023 The University of Hong Kong, Department of Computer Science
# All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are met:
#
# 1. Redistributions of source code must retain the above copyright notice, this
#    list of conditions and the following disclaimer.
#
# 2. Redistributions in binary form must reproduce the above copyright notice,
#    this list of conditions and the following disclaimer in the documentation
#    and/or other materials provided with the distribution.
#
# 3. Neither the name of the copyright holder nor the names of its
#    contributors may be used to endorse or promote products derived from
#    this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
# AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
# DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
# SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
# CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
# OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

import os
import sys
import argparse
import shlex
import subprocess

from collections import defaultdict, namedtuple
from argparse import SUPPRESS
try:
    from packaging.version import parse as version_parse
except ModuleNotFoundError:
    from distutils.version import LooseVersion as version_parse

from time import time

import shared.param as param
from shared.interval_tree import bed_tree_from
from shared.utils import file_path_from, folder_path_from, subprocess_popen, str2bool, str_none, \
    legal_range_from, log_error, log_warning, clair3_option_type


major_contigs = {"chr" + str(a) for a in list(range(1, 23)) + ["X", "Y"]}.union(
    {str(a) for a in list(range(1, 23)) + ["X", "Y"]})
major_contigs_order = ["chr" + str(a) for a in list(range(1, 23)) + ["X", "Y"]] + [str(a) for a in
                                                                                   list(range(1, 23)) + ["X", "Y"]]


file_directory = os.path.dirname(os.path.realpath(__file__))
main_entry = os.path.join(file_directory, "clairs.py")
MAX_STEP = 20

OutputPath = namedtuple('OutputPath', [
    'log_path',
    'tmp_file_path',
    'split_bed_path',
    'candidates_path',
    'pileup_tensor_can_path',
    'fa_tensor_can_path',
    'vcf_output_path',
    'tmp_vcf_output_path',
])

Clair3Option = namedtuple('Clair3Option', [
    'clair3_path',
    'model_path',
    'longphase_for_phasing',
    'ctg_name_str',
    'min_coverage',
    'snp_min_af',
    'indel_min_af',
    'longphase'
])

supported_platforms = ['ont_r10_dorado_sup_5khz_ssrs', 'ont_r10_dorado_sup_5khz_ss', 'ont_r10_dorado_sup_5khz', 'ont_r10_dorado_sup_4khz', 'ont_r10_dorado_hac_5khz',
                       'ont_r10_dorado_hac_4khz', 'ont_r10_guppy', 'ont_r9_guppy', 'ilmn', 'hifi_sequel2', 'hifi_revio', 'hifi_revio_ssrs', 'hifi_revio_ss','ilmn_ssrs', 'ilmn_ss'
                       ]

indel_supported_platforms = ['ont_r10_dorado_sup_5khz_ssrs', 'ont_r10_dorado_sup_5khz_ss', 'ont_r10_guppy', 'ont_r10', 'ont', 'hifi_revio', 'hifi_sequel2', 'ont_r10_dorado_sup_5khz',
                       'ont_r10_dorado_hac_5khz', 'ont_r10_dorado_sup_4khz', 'ont_r10_dorado_hac_5khz_liquid', 'hifi_revio_ssrs', 'hifi_revio_ss']

extend_supported_platforms = supported_platforms + ['ont_r10', 'ont_r9', 'ont', 'ont_r10_dorado_hac_5khz_liquid']


class Tee(object):
    def __init__(self, name, mode):
        self.file = open(name, mode)
        self.stdout = sys.stdout
        sys.stdout = self

    def __del__(self):
        sys.stdout = self.stdout
        self.file.close()

    def write(self, data):
        self.file.write(data)
        self.stdout.write(data)

    def flush(self):
        self.file.flush()

def logging(str):
    if args.tee is None:
        print(str)
    else:
        args.tee.stdin.write(bytes(str + '\n', encoding='utf8'))

def create_output_folder(args):
    # create temp file folder
    args.output_dir = folder_path_from(args.output_dir, create_not_found=True)
    log_path = folder_path_from(os.path.join(args.output_dir, 'logs'), create_not_found=True)
    tmp_file_path = folder_path_from(os.path.join(args.output_dir, 'tmp'), create_not_found=True)
    split_bed_path = folder_path_from(os.path.join(tmp_file_path, 'split_beds'), create_not_found=True)
    candidates_path = folder_path_from(os.path.join(tmp_file_path, 'candidates'), create_not_found=True)
    pileup_tensor_can_path = folder_path_from(os.path.join(tmp_file_path, 'pileup_tensor_can'), create_not_found=True)
    fa_tensor_can_path = folder_path_from(os.path.join(tmp_file_path, 'fa_tensor_can'), create_not_found=True)
    vcf_output_path = folder_path_from(os.path.join(tmp_file_path, 'vcf_output'), create_not_found=True)
    tmp_vcf_output_path = folder_path_from(os.path.join(tmp_file_path, 'tmp_vcf_output'), create_not_found=True)

    if args.platform != 'ilmn':
        clair3_log_path = folder_path_from(os.path.join(args.output_dir, 'logs', 'clair3_log'), create_not_found=True)
        clair3_phased_output_path = folder_path_from(os.path.join(tmp_file_path, 'clair3_output/phased_output'), create_not_found=True)
        clair3_vcf_output_path = folder_path_from(os.path.join(tmp_file_path, 'clair3_output/vcf'), create_not_found=True)

    output_path = OutputPath(log_path=log_path,
                             tmp_file_path=tmp_file_path,
                             split_bed_path=split_bed_path,
                             candidates_path=candidates_path,
                             pileup_tensor_can_path=pileup_tensor_can_path,
                             fa_tensor_can_path=fa_tensor_can_path,
                             tmp_vcf_output_path=tmp_vcf_output_path,
                             vcf_output_path=vcf_output_path)
    return output_path

def check_version(tool, pos=None, is_pypy=False):
    try:
        if is_pypy:
            proc = subprocess.run("{} -c 'import sys; print (sys.version)'".format(tool), stdout=subprocess.PIPE,
                                  shell=True)
        else:
            proc = subprocess.run([tool, "--version"], stdout=subprocess.PIPE)
        if proc.returncode != 0:
            return None
        first_line = proc.stdout.decode().split("\n", 1)[0]
        version = first_line.split()[pos]
        version = version_parse(version)
    except Exception:
        return None

    return version

def check_skip_steps_legal(args):
    skip_steps = args.skip_steps
    skip_steps_list = skip_steps.rstrip().split(",")
    if len(skip_steps_list) == 0:
        sys.exit(log_error("[ERROR] --skip_steps option provided but no skip steps index found"))
    for step in skip_steps_list:
        if int(step) < 1 or int(step) > MAX_STEP:
            sys.exit(log_error("[ERROR] --skip_steps option provided but contains invalid skip steps index, should be 1-index"))

def check_python_path():
    python_path = subprocess.run("which python", stdout=subprocess.PIPE, shell=True).stdout.decode().rstrip()
    sys.exit(log_error("[ERROR] Current python execution path: {}".format(python_path)))

def check_python_version(python):
    python_path = subprocess.run("{} --version".format(python), stdout=subprocess.PIPE, shell=True).stdout.decode().rstrip()
    return python_path.split(' ')[1]


def check_tools_version(args):

    required_tool_version = {
        'python': version_parse('3.9.0'),
        'pypy': version_parse('3.6'),
        'samtools': version_parse('1.10'),
        'whatshap': version_parse('1.0'),
        'parallel': version_parse('20191122'),
    }

    tool_version = {
        'python': version_parse(check_python_version(args.python)),
        'pypy': check_version(tool=args.pypy, pos=0, is_pypy=True),
        'samtools': check_version(tool=args.samtools, pos=1),
        'parallel': check_version(tool=args.parallel, pos=2),
    }

    for tool, version in tool_version.items():
        required_version = required_tool_version[tool]
        if version is None:
            logging(log_error("[ERROR] {} not found, please check if you are in the conda virtual environment".format(tool)))
            check_python_path()
        elif version < required_version:
            logging(log_error("[ERROR] Tool version not match, please check if you are in the conda virtual environment"))
            logging(' '.join([str(item).ljust(10) for item in ["Tool", "Version", "Required"]]))
            error_info = ' '.join([str(item).ljust(10) for item in [tool, version, '>=' + str(required_version)]])
            logging(error_info)
            check_python_path()
    return


def check_contig_in_bam(bam_fn, sorted_contig_list, samtools, allow_none=False, is_tumor=False):
    flag = 'tumor' if is_tumor else 'normal'
    if allow_none and bam_fn is None:
        return sorted_contig_list, True
    bai_process = subprocess_popen(shlex.split("{} idxstats {}".format(samtools, bam_fn)))
    contig_with_read_support_set = set()
    for row_id, row in enumerate(bai_process.stdout):
        row = row.split('\t')
        if len(row) != 4:
            continue
        contig_name, contig_length, mapped_reads, unmapped_reads = row
        if contig_name not in sorted_contig_list:
            continue
        if int(mapped_reads) > 0:
            contig_with_read_support_set.add(contig_name)
    for contig_name in sorted_contig_list:
        if contig_name not in contig_with_read_support_set:
            logging(log_warning(
                "[WARNING] Contig name {} provided but no mapped reads found in {} BAM, skip!".format(contig_name, flag)))
    filtered_sorted_contig_list = [item for item in sorted_contig_list if item in contig_with_read_support_set]

    found_contig = True
    if len(filtered_sorted_contig_list) == 0:
        found_contig = False
        logging(log_warning(
            "[WARNING] No mapped reads found in {} BAM for provided contigs set {}".format(
                flag, ' '.join(sorted_contig_list))))

    return filtered_sorted_contig_list, found_contig


def check_threads(args):
    threads = args.threads
    #sched_getaffinity is not exist in pypy
    try:
        sched_getaffinity_list = list(os.sched_getaffinity(0))
        num_cpus = len(sched_getaffinity_list)
    except:
        num_cpus = int(subprocess.run(args.python + " -c \"import os; print(len(os.sched_getaffinity(0)))\"", \
                                      stdout=subprocess.PIPE, shell=True).stdout.decode().rstrip())

    if threads > num_cpus:
        logging(log_warning(
            '[WARNING] Threads setting {} is larger than the number of available threads {} in the system,'.format(
                threads, num_cpus)))
        logging(log_warning('Set --threads={} for better parallelism.'.format(num_cpus)))
        args.threads = num_cpus
    return args

def split_extend_vcf(genotyping_mode_vcf_fn, output_fn):
    expand_region_size = param.no_of_positions
    output_ctg_dict = defaultdict(list)
    unzip_process = subprocess_popen(shlex.split("gzip -fdc %s" % (genotyping_mode_vcf_fn)))

    for row_id, row in enumerate(unzip_process.stdout):
        if row[0] == '#':
            continue
        columns = row.strip().split(maxsplit=3)
        ctg_name = columns[0]

        center_pos = int(columns[1])
        ctg_start, ctg_end = center_pos - 1, center_pos
        if ctg_start < 0:
            sys.exit(
                log_error("[ERROR] Invalid VCF input at the {}-th row {} {} {}".format(row_id + 1, ctg_name, center_pos)))
        if ctg_start - expand_region_size < 0:
            continue
        expand_ctg_start = ctg_start - expand_region_size
        expand_ctg_end = ctg_end + expand_region_size

        output_ctg_dict[ctg_name].append(
            ' '.join([ctg_name, str(expand_ctg_start), str(expand_ctg_end)]))

    for key, value in output_ctg_dict.items():
        ctg_output_fn = os.path.join(output_fn, key)
        with open(ctg_output_fn, 'w') as output_file:
            output_file.write('\n'.join(value))

    unzip_process.stdout.close()
    unzip_process.wait()

    know_vcf_contig_set = set(list(output_ctg_dict.keys()))

    return know_vcf_contig_set


def split_extend_bed(bed_fn, output_fn, contig_set=None):
    expand_region_size = param.no_of_positions
    output_ctg_dict = defaultdict(list)
    unzip_process = subprocess_popen(shlex.split("gzip -fdc %s" % (bed_fn)))
    for row_id, row in enumerate(unzip_process.stdout):
        if row[0] == '#':
            continue
        columns = row.strip().split()
        ctg_name = columns[0]
        if contig_set and ctg_name not in contig_set:
            continue

        ctg_start, ctg_end = int(columns[1]), int(columns[2])

        if ctg_end < ctg_start or ctg_start < 0 or ctg_end < 0:
            sys.exit(log_error(
                "[ERROR] Invalid BED input at the {}-th row {} {} {}".format(row_id + 1, ctg_name, ctg_start, ctg_end)))
        expand_ctg_start = max(0, ctg_start - expand_region_size)
        expand_ctg_end = max(0, ctg_end + expand_region_size)
        output_ctg_dict[ctg_name].append(
            ' '.join([ctg_name, str(expand_ctg_start), str(expand_ctg_end)]))

    for key, value in output_ctg_dict.items():
        ctg_output_fn = os.path.join(output_fn, key)
        with open(ctg_output_fn, 'w') as output_file:
            output_file.write('\n'.join(value))

    unzip_process.stdout.close()
    unzip_process.wait()

def write_region_bed(region):

    try:
        ctg_name, start_end = region.split(':')
        ctg_start, ctg_end = int(start_end.split('-')[0]) - 1, int(start_end.split('-')[1]) - 1  # bed format
    except:
        sys.exit("[ERROR] Please use the correct format for --region: ctg_name:start-end, your input is {}".format(
            region))
    if ctg_end < ctg_start or ctg_start < 0 or ctg_end < 0:
        sys.exit("[ERROR] Invalid region input: {}".format(region))

    output_bed_path = os.path.join(args.output_dir, 'tmp', 'region.bed')
    with open(output_bed_path, 'w') as f:
        f.write('\t'.join([ctg_name, str(ctg_start), str(ctg_end)]) + '\n')
    return output_bed_path

def check_contigs_intersection(args, fai_fn):

    MIN_CHUNK_LENGTH = 200000
    MAX_CHUNK_LENGTH = 20000000
    is_include_all_contigs = args.include_all_ctgs
    is_bed_file_provided = args.bed_fn is not None or args.region is not None
    is_known_vcf_file_provided = args.genotyping_mode_vcf_fn is not None
    is_ctg_name_list_provided = args.ctg_name is not None

    if args.region is not None:
        args.bed_fn = write_region_bed(args.region)

    split_bed_path = os.path.join(args.output_dir, 'tmp', 'split_beds')
    tree = bed_tree_from(bed_file_path=args.bed_fn, region=args.region)
    know_vcf_contig_set = split_extend_vcf(genotyping_mode_vcf_fn=args.genotyping_mode_vcf_fn, output_fn=split_bed_path) if is_known_vcf_file_provided else set()
    contig_set = set(args.ctg_name.split(',')) if is_ctg_name_list_provided else set()

    if not args.include_all_ctgs:
        logging("[INFO] --include_all_ctgs not enabled, use chr{1..22,X,Y} and {1..22,X,Y} by default")
    else:
        logging("[INFO] --include_all_ctgs enabled")

    if is_ctg_name_list_provided and is_bed_file_provided:
        logging(log_warning("[WARNING] both --ctg_name and --bed_fn provided, will only proceed with the contigs appeared in both"))

    if is_ctg_name_list_provided and is_known_vcf_file_provided:
        logging(log_warning("[WARNING] both --ctg_name and --genotyping_mode_vcf_fn provided, will only proceed with the contigs appeared in both"))

    if is_ctg_name_list_provided:
        contig_set = contig_set.intersection(
            set(tree.keys())) if is_bed_file_provided else contig_set
        contig_set = contig_set.intersection(
            know_vcf_contig_set) if is_known_vcf_file_provided else contig_set
    else:
        contig_set = contig_set.union(
            set(tree.keys())) if is_bed_file_provided else contig_set

        contig_set = contig_set.union(
            know_vcf_contig_set) if is_known_vcf_file_provided else contig_set

    # if each split region is too small(long) for given default chunk num, will increase(decrease) the total chunk num
    default_chunk_num = 0
    DEFAULT_CHUNK_SIZE = args.chunk_size
    contig_length_list = []
    contig_chunk_num = {}

    with open(fai_fn, 'r') as fai_fp:
        for row in fai_fp:
            columns = row.strip().split("\t")
            contig_name, contig_length = columns[0], int(columns[1])
            if not is_include_all_contigs and (
            not (is_bed_file_provided or is_ctg_name_list_provided or is_known_vcf_file_provided)) and str(
                    contig_name) not in major_contigs:
                continue

            if is_bed_file_provided and contig_name not in tree:
                continue
            if is_ctg_name_list_provided and contig_name not in contig_set:
                continue
            if is_known_vcf_file_provided and contig_name not in contig_set:
                continue

            contig_set.add(contig_name)
            contig_length_list.append(contig_length)
            chunk_num = int(
                contig_length / float(DEFAULT_CHUNK_SIZE)) + 1 if contig_length % DEFAULT_CHUNK_SIZE else int(
                contig_length / float(DEFAULT_CHUNK_SIZE))
            contig_chunk_num[contig_name] = max(chunk_num, 1)

    if default_chunk_num > 0:
        min_chunk_length = min(contig_length_list) / float(default_chunk_num)
        max_chunk_length = max(contig_length_list) / float(default_chunk_num)

    contigs_order = major_contigs_order + list(contig_set)

    sorted_contig_list = sorted(list(contig_set), key=lambda x: contigs_order.index(x))

    found_contig = True
    if not len(contig_set):
        if is_bed_file_provided:
            all_contig_in_bed = ' '.join(list(tree.keys()))
            logging(log_warning("[WARNING] No contig in --bed_fn was found in the reference, contigs in BED {}: {}".format(args.bed_fn, all_contig_in_bed)))
        if is_known_vcf_file_provided:
            all_contig_in_vcf = ' '.join(list(know_vcf_contig_set))
            logging(log_warning("[WARNING] No contig in --genotyping_mode_vcf_fn was found in the reference, contigs in VCF {}: {}".format(args.genotyping_mode_vcf_fn, all_contig_in_vcf)))
        if is_ctg_name_list_provided:
            all_contig_in_ctg_name = ' '.join(args.ctg_name.split(','))
            logging(log_warning("[WARNING] No contig in --ctg_name was found in the reference, contigs in contigs list: {}".format(all_contig_in_ctg_name)))
        found_contig = False
    else:
        for c in sorted_contig_list:
            if c not in contig_chunk_num:
                logging(log_warning(("[WARNING] Contig {} given but not found in the reference".format(c))))

        # check contig in bam have support reads
        sorted_contig_list, tumor_found_contig = check_contig_in_bam(bam_fn=args.tumor_bam_fn, sorted_contig_list=sorted_contig_list,
                                                               samtools=args.samtools, is_tumor=True)

        sorted_contig_list, normal_found_contig = check_contig_in_bam(bam_fn=args.normal_bam_fn, sorted_contig_list=sorted_contig_list,
                                                               samtools=args.samtools, allow_none=False, is_tumor=False)
        found_contig = tumor_found_contig and normal_found_contig

    if not found_contig:
        log_warning("[WARNING] Exit calling because no contig was found in BAM!")
        sys.exit(0)
    logging('[INFO] Call variants in contigs: {}'.format(' '.join(sorted_contig_list)))
    logging('[INFO] Number of chunks for each contig: {}'.format(
        ' '.join([str(contig_chunk_num[c]) for c in sorted_contig_list])))

    if default_chunk_num > 0 and max_chunk_length > MAX_CHUNK_LENGTH:
        logging(log_warning(
            '[WARNING] The maximum chunk size set {} is larger than the suggested maximum chunk size {}, consider setting a larger --chunk_num= instead for better parallelism.'.format(
                min_chunk_length, MAX_CHUNK_LENGTH)))

    elif default_chunk_num > 0 and min_chunk_length < MIN_CHUNK_LENGTH:
        logging(log_warning(
            '[WARNING] The minimum chunk size set {} is smaller than the suggested  minimum chunk size {}, consider setting a smaller --chunk_num= instead.'.format(
                min_chunk_length, MIN_CHUNK_LENGTH)))

    if default_chunk_num == 0 and max(contig_length_list) < DEFAULT_CHUNK_SIZE / 5:
        logging(log_warning(
            '[WARNING] The length of the longest contig {} is more than five times smaller smaller than the default chunk size {}, consider setting a smaller --chunk_size= instead for better parallelism.'.format(
                max(contig_length_list), DEFAULT_CHUNK_SIZE)))

    if is_bed_file_provided:
        split_extend_bed(bed_fn=args.bed_fn, output_fn=split_bed_path, contig_set=contig_set)

    contig_path = os.path.join(args.output_dir, 'tmp', 'CONTIGS')
    with open(contig_path, 'w') as output_file:
        output_file.write('\n'.join(sorted_contig_list))

    chunk_list = []
    chunk_list_path = os.path.join(args.output_dir, 'tmp', 'CHUNK_LIST')
    with open(chunk_list_path, 'w') as output_file:
        for contig_name in sorted_contig_list:
            chunk_num = contig_chunk_num[contig_name] if args.chunk_num is None else args.chunk_num
            for chunk_id in range(1, chunk_num + 1):
                output_file.write(contig_name + ' ' + str(chunk_id) + ' ' + str(chunk_num) + '\n')
                chunk_list.append((contig_name, chunk_id, chunk_num))
    args.chunk_list = chunk_list
    if args.clair3_path is not None and args.platform != 'ilmn':
        args.clair3_option = args.clair3_option._replace(ctg_name_str=','.join(sorted_contig_list))

    return args

def check_clair3_options(args):
    conda_prefix = args.conda_prefix
    if args.clair3_path is None:
        args.clair3_path = os.path.join(conda_prefix, 'bin')
        if not os.path.exists(args.clair3_path + '/run_clair3.sh'):
            sys.exit(log_error("[ERROR] Cannot find clair3 main entry in {}".format(args.clair3_path)))

    if args.clair3_model_path is None:
        if args.platform == 'ont_r10' or args.platform == 'ont' or args.platform == 'ont_r10_guppy':
            args.clair3_model_path = os.path.join(conda_prefix, 'bin', 'clairs_models', 'clair3_models', 'ont_r104_e81_sup_g5015')
        elif args.platform == 'ont_r9' or args.platform == 'ont_r9_guppy':
            # clair3 r9 model in bioconda bin
            args.clair3_model_path = os.path.join(conda_prefix, 'bin', 'models', 'r941_prom_sup_g5014')
        elif args.platform == 'hifi' or args.platform == 'hifi_sequel2':
            # clair3 hifi model in bioconda bin
            args.clair3_model_path = os.path.join(conda_prefix, 'bin', 'models', 'hifi')
        elif args.platform.startswith("hifi_revio"):
            args.clair3_model_path = os.path.join(conda_prefix, 'bin', 'clairs_models', 'clair3_models', 'hifi_revio')
        elif args.platform == "ont_r10_dorado_4khz" or args.platform == "ont_r10_dorado_sup_4khz":
            args.clair3_model_path = os.path.join(conda_prefix, 'bin', 'clairs_models', 'clair3_models', 'r1041_e82_400bps_sup_v410')
        elif args.platform == "ont_r10_dorado_5khz" or args.platform.startswith("ont_r10_dorado_sup_5khz"):
            args.clair3_model_path = os.path.join(conda_prefix, 'bin', 'clairs_models', 'clair3_models', 'r1041_e82_400bps_sup_v420')
        elif args.platform == "ont_r10_dorado_hac_5khz" or args.platform == "ont_r10_dorado_hac_5khz_liquid":
            args.clair3_model_path = os.path.join(conda_prefix, 'bin', 'clairs_models', 'clair3_models', 'r1041_e82_400bps_hac_v420')
        elif args.platform == "ont_r10_dorado_hac_4khz":
            args.clair3_model_path = os.path.join(conda_prefix, 'bin', 'clairs_models', 'clair3_models', 'r1041_e82_400bps_hac_v410')

    if not os.path.exists(args.clair3_model_path):
        sys.exit(log_error("[ERROR] Cannot find clair3 model path in {}".format(args.clair3_model_path)))

    if args.longphase is None:
        args.longphase = os.path.join(conda_prefix, 'bin', 'longphase')
        if not os.path.exists(args.longphase):
            args.longphase = os.path.join(args.clair3_path, 'longphase')

    if args.whatshap is None:
        args.whatshap = os.path.join(conda_prefix, 'bin', 'whatshap')

    longphase_for_phasing = False if args.whatshap_for_phasing is True else param.clair3_fast_option['longphase_for_phasing']
    if longphase_for_phasing and not os.path.exists(args.longphase):
        sys.exit(log_error("[ERROR] Cannot find longphase at {}".format(args.longphase)))
    if args.whatshap_for_phasing and not os.path.exists(args.whatshap):
        sys.exit(log_error("[ERROR] Cannot find whatshap at {}".format(args.whatshap)))

    clair3_option = param.clair3_fast_option
    if args.enable_clair3_germline_output:
        # use clair3 default settings than the fast setting for germline calling
        clair3_option = param.clair3_default_option

    args.clair3_min_coverage = args.clair3_min_coverage if args.clair3_min_coverage is not None else clair3_option['min_coverage']
    args.clair3_snp_min_af = args.clair3_snp_min_af if args.clair3_snp_min_af is not None else clair3_option['snp_min_af']
    args.clair3_indel_min_af = args.clair3_indel_min_af if args.clair3_indel_min_af is not None else clair3_option['indel_min_af']
    args.clair3_option = Clair3Option(
        clair3_path=args.clair3_path,
        model_path=args.clair3_model_path,
        longphase_for_phasing=longphase_for_phasing,
        ctg_name_str=None, # update in check_contigs_intersection
        min_coverage=str(args.clair3_min_coverage),
        snp_min_af=str(args.clair3_snp_min_af),
        indel_min_af=str(args.clair3_indel_min_af),
        longphase=args.longphase)

    return args

def check_args(args):

    if args.conda_prefix is None:
        if 'CONDA_PREFIX' in os.environ:
            args.conda_prefix = os.environ['CONDA_PREFIX']
        else:
            try:
                python_path = subprocess.run('which python', stdout=subprocess.PIPE, shell=True).stdout.decode().rstrip()
                args.conda_prefix = os.path.dirname(os.path.dirname(python_path))
            except:
                sys.exit(log_error("[ERROR] Conda prefix not found, please activate a correct conda environment."))

    args.normal_bam_fn = file_path_from(file_name=args.normal_bam_fn, exit_on_not_found=True, allow_none=False)
    normal_bai_fn = file_path_from(file_name=args.normal_bam_fn, suffix=".bai", exit_on_not_found=False, sep='.')
    normal_crai_fn = file_path_from(file_name=args.normal_bam_fn, suffix=".crai", exit_on_not_found=False, sep='.')
    normal_csi_fn = file_path_from(file_name=args.normal_bam_fn, suffix=".csi", exit_on_not_found=False, sep='.')

    args.tumor_bam_fn = file_path_from(file_name=args.tumor_bam_fn, exit_on_not_found=True)
    tumor_bai_fn = file_path_from(file_name=args.tumor_bam_fn, suffix=".bai", exit_on_not_found=False, sep='.')
    tumor_crai_fn = file_path_from(file_name=args.tumor_bam_fn, suffix=".crai", exit_on_not_found=False, sep='.')
    tumor_csi_fn = file_path_from(file_name=args.tumor_bam_fn, suffix=".csi", exit_on_not_found=False, sep='.')

    args.ref_fn = file_path_from(file_name=args.ref_fn, exit_on_not_found=True)
    fai_fn = file_path_from(file_name=args.ref_fn, suffix=".fai", exit_on_not_found=True, sep='.')
    args.bed_fn = file_path_from(file_name=args.bed_fn, exit_on_not_found=True, allow_none=True)
    args.genotyping_mode_vcf_fn = file_path_from(file_name=args.genotyping_mode_vcf_fn, exit_on_not_found=True, allow_none=True)
    args.hybrid_mode_vcf_fn = file_path_from(file_name=args.hybrid_mode_vcf_fn, exit_on_not_found=True, allow_none=True)
    args.normal_vcf_fn = file_path_from(file_name=args.normal_vcf_fn, exit_on_not_found=True, allow_none=True)

    if args.platform in param.model_name_platform_dict:
        updated_platform = param.model_name_platform_dict[args.platform]
        logging("[INFO] Platform parameter is using ONT model name format, change --platform {} to --platform {}".format(args.platform, updated_platform))
        args.platform = updated_platform

    if normal_bai_fn is None and normal_crai_fn is None and normal_csi_fn is None:
        sys.exit(log_error("[ERROR] Normal BAM index file {} or {} not found. Please run `samtools index $BAM` first.".format(args.normal_bam_fn + '.bai',
                                                                                      args.normal_bam_fn + '.crai')))

    if tumor_bai_fn is None and tumor_crai_fn is None and tumor_csi_fn is None:
        sys.exit(log_error("[ERROR] Tumor BAM index file {} or {} not found. Please run `samtools index $BAM` first.".format(args.tumor_bam_fn + '.bai',
                                                                                      args.tumor_bam_fn + '.crai')))

    if args.enable_indel_calling and args.platform not in set(indel_supported_platforms):
        sys.exit(log_error("[ERROR] Indel calling only support {} platforms".format(','.join(indel_supported_platforms))))

    if args.genotyping_mode_vcf_fn is not None and args.hybrid_mode_vcf_fn is not None:
        sys.exit(log_error("[ERROR] Please provide either --genotyping_mode_vcf_fn or --hybrid_mode_vcf_fn only"))


    if args.platform == 'ont_r10_dorado_hac_5khz_liquid':
        args.enable_params_for_liquid_tumor_sample = True

    if args.pileup_model_path is None:
        args.pileup_model_path = os.path.join(args.conda_prefix, 'bin', 'clairs_models', args.platform, 'pileup.pkl')

    if args.full_alignment_model_path is None:
        args.full_alignment_model_path = os.path.join(args.conda_prefix, 'bin', 'clairs_models', args.platform, 'full_alignment.pkl')

    args.pileup_model_path = file_path_from(file_name=args.pileup_model_path, exit_on_not_found=True, is_directory=False, allow_none=False)
    args.full_alignment_model_path = file_path_from(file_name=args.full_alignment_model_path, exit_on_not_found=True, is_directory=False, allow_none=False)

    if args.enable_indel_calling:
        if args.indel_pileup_model_path is None:
            args.indel_pileup_model_path = os.path.join(args.conda_prefix, 'bin', 'clairs_models', args.platform, 'indel',
                                                  'pileup.pkl')

        if args.indel_full_alignment_model_path is None:
            args.indel_full_alignment_model_path = os.path.join(args.conda_prefix, 'bin', 'clairs_models', args.platform, 'indel',
                                                          'full_alignment.pkl')

        args.indel_pileup_model_path = file_path_from(file_name=args.indel_pileup_model_path, exit_on_not_found=True,
                                                is_directory=False, allow_none=False)
        args.indel_full_alignment_model_path = file_path_from(file_name=args.indel_full_alignment_model_path,
                                                        exit_on_not_found=True, is_directory=False, allow_none=False)

    if args.snv_min_af is None:
        args.snv_min_af = param.snv_min_af
    if args.indel_min_af is None:
        if args.enable_indel_calling:
            if args.platform.startswith('ont'):
                args.indel_min_af = 0.1
            else:
                args.indel_min_af = 0.05
        else:
            args.indel_min_af = 1.0
    if args.min_coverage is None:
        args.min_coverage = param.min_coverage
    if args.chunk_size is None:
        args.chunk_size = 5000000
    if args.platform not in set(extend_supported_platforms):
        logging(log_error('[ERROR] Invalid platform input, optional: {}'.format(', '.join(supported_platforms))))

    if (args.qual is not None or args.prefer_balance or args.prefer_recall) and (args.snv_min_qual is not None or args.indel_min_qual is not None):
        option = '--qual' if args.qual is not None else "--prefer_balance" if args.prefer_balance else "--prefer_recall"
        if args.snv_min_qual is not None:
            logging(log_warning(
                f"[WARNING] {option} is set, but --snv_min_qual is also set, will use --snv_min_qual instead of {option} for SNV calling."))
        if args.indel_min_qual is not None:
            logging(log_warning(
                f"[WARNING] {option} is set, but --indel_min_qual is also set, will use --indel_min_qual instead of {option} for Indel calling."))
        args.qual = None
        args.prefer_balance = False

    if args.qual is None:
        if args.prefer_balance:
            args.qual = param.best_thred_qual[args.platform] if args.platform in param.best_thred_qual else param.best_thred_qual['ont']
            args.snv_min_qual = args.indel_min_qual = args.qual
        else:
            # default we enable min_thred_qual
            if args.snv_min_qual is None and args.indel_min_qual is None:
                args.qual = param.min_thred_qual[args.platform] if args.platform in param.min_thred_qual else param.min_thred_qual['ont']
                args.snv_min_qual = args.indel_min_qual = args.qual
            elif args.snv_min_qual is None:
                args.qual = args.snv_min_qual = param.min_thred_qual[args.platform] if args.platform in param.min_thred_qual else param.min_thred_qual['ont']
            elif args.indel_min_qual is None:
                args.qual = args.indel_min_qual = param.min_thred_qual[args.platform] if args.platform in param.min_thred_qual else param.min_thred_qual['ont']

    model_path = os.path.dirname(args.pileup_model_path)
    config_file = os.path.join(model_path, 'model_specific_settings.conf')
    if os.path.exists(config_file):
        with open(config_file, 'r') as config_fp:
            for line in config_fp:
                if line.startswith('snv_min_qual'):
                    args.snv_min_qual = int(line.split('=')[1].strip())
                    print("[INFO] Set SNV minimum quality `--snv_min_qual` to {} according to the parameter config file".format(args.snv_min_qual))
                elif line.startswith('indel_min_qual'):
                    args.indel_min_qual = int(line.split('=')[1].strip())
                    print("[INFO] Set Indel minimum quality `--indel_min_qual` to {} according to the parameter config file".format(args.indel_min_qual))

    if args.skip_steps is not None:
        check_skip_steps_legal(args)
    if args.enable_realignment and args.platform != 'ilmn':
        args.enable_realignment = False

    if args.min_bq is None and (args.platform == "ont_r10_dorado_hac_5khz" or args.platform == "ont_r10_dorado_hac_4khz" or args.platform == 'ont_r10_dorado_hac_5khz_liquid'):
        args.min_bq = 15

    if args.disable_phasing or args.haplotagged_tumor_bam_provided_so_skip_intermediate_phasing_and_haplotagging:
        args.phase_tumor = False
    if args.phase_tumor is None:
        if args.genotyping_mode_vcf_fn is not None:
            logging(log_warning(
                "[WARNING] Clair3 HET SNPs based phasing is disabled if `--genotyping_mode_vcf_fn` is provided, add `--phase_tumor True` if phasing the tumor is still needed. Please ensure you have sufficient heterozygous variant candidates given in the --genotyping_mode_vcf_fn file, otherwise the phasing step might lead to worse performance."))
        else:
            args.phase_tumor = True if args.platform != 'ilmn' else False

    #N: normal V: variant T: tumor B: BAM
    NVNB = args.use_heterozygous_snp_in_normal_sample_and_normal_bam_for_intermediate_phasing
    TVTB = args.use_heterozygous_snp_in_tumor_sample_and_tumor_bam_for_intermediate_phasing
    NVTB = args.use_heterozygous_snp_in_normal_sample_and_tumor_bam_for_intermediate_phasing
    TVNB = args.use_heterozygous_snp_in_tumor_sample_and_normal_bam_for_intermediate_phasing

    # By default, we use HET SNP in normal VCF for phasing after v0.1.7
    if NVNB is None and TVTB is None and NVTB is None and TVNB is None:
        args.use_heterozygous_snp_in_normal_sample_and_tumor_bam_for_intermediate_phasing = True
        NVTB = True
    if NVTB:
        args.use_heterozygous_snp_in_normal_sample_for_intermediate_phasing = True
        args.use_normal_bam_for_intermediate_phasing = False
    elif TVNB:
        args.use_heterozygous_snp_in_tumor_sample_for_intermediate_phasing = True
        args.use_normal_bam_for_intermediate_phasing = True
    elif NVNB:
        args.use_heterozygous_snp_in_normal_sample_for_intermediate_phasing = True
        args.use_normal_bam_for_intermediate_phasing = True
    elif TVTB:
        args.use_heterozygous_snp_in_tumor_sample_for_intermediate_phasing = True
        args.use_normal_bam_for_intermediate_phasing = False

    if args.use_heterozygous_indel_for_intermediate_phasing is None:
        args.use_heterozygous_indel_for_intermediate_phasing = False

    if args.use_longphase_for_intermediate_haplotagging is None:
        args.use_longphase_for_intermediate_haplotagging = False

    if args.enable_verdict:
        ref_contigs_set = set()
        with open(fai_fn, 'r') as fai_fp:
            for row in fai_fp:
                columns = row.strip().split("\t")
                contig_name = columns[0]
                ref_contigs_set.add(contig_name)
        contigs_order = ["chr" + str(a) for a in list(range(1, 23)) + ["X"]]
        verdict_flag = set(contigs_order).intersection(ref_contigs_set) is not None
        if not verdict_flag:
            args.enable_verdict = False
            logging(log_warning(
                "[WARNING] Verdict currently only works for GRCh38 reference genome, disable the --enable_verdict option!"))
        if args.cnv_resource_dir is None:
            args.cnv_resource_dir = os.path.join(args.conda_prefix, 'bin', 'cnv_data', 'reference_files')
        if args.allele_counter_dir is None:
            args.allele_counter_dir = os.path.join(file_directory, 'src', 'verdict', 'allele_counter')
        if not os.path.exists(args.allele_counter_dir):
            args.enable_verdict = False
            logging(log_warning(
                "[WARNING] The allele counter {}is not found, disable the --enable_verdict option!".format(args.allele_counter_dir)))
        if not os.path.exists(args.cnv_resource_dir):
            args.enable_verdict = False
            logging(log_warning(
                "[WARNING] The CNV resource directory {} is not found, disable the --enable_verdict option!".format(args.cnv_resource_dir)))

    if args.genotyping_mode_vcf_fn is not None or args.hybrid_mode_vcf_fn is not None:
        logging(log_warning("[INFO] Enable --print_ref_calls and --print_germline_calls options in genotyping mode!"))
        args.print_ref_calls = True
        args.print_germline_calls = True
        if args.hybrid_mode_vcf_fn is not None and args.platform != 'ilmn':
            logging(log_warning(
                "[INFO] Enable --enable_clair3_germline_output option in hybrid calling mode!"))
            args.enable_clair3_germline_output = True

    if args.phase_tumor is False:
        # cannot only phase normal
        args.phase_normal = False

    legal_range_from(param_name="threads", x=args.threads, min_num=1, exit_out_of_range=True)
    legal_range_from(param_name="snv_min_qual", x=args.snv_min_qual, min_num=-1, exit_out_of_range=True)
    legal_range_from(param_name="indel_min_qual", x=args.indel_min_qual, min_num=-1, exit_out_of_range=True)
    legal_range_from(param_name="min_coverage", x=args.min_coverage, min_num=0, exit_out_of_range=True)
    legal_range_from(param_name="snv_min_af", x=args.snv_min_af, min_num=0, max_num=1, exit_out_of_range=True)
    legal_range_from(param_name="indel_min_af", x=args.indel_min_af, min_num=0, max_num=1, exit_out_of_range=True)
    if args.indel_max_af_in_normal is not None:
        legal_range_from(param_name="indel_max_af_in_normal", x=args.indel_max_af_in_normal, min_num=0, max_num=1, exit_out_of_range=True)
    legal_range_from(param_name="chunk_size", x=args.chunk_size, min_num=0, exit_out_of_range=True)

    args.output_path = create_output_folder(args)
    check_tools_version(args=args)
    args = check_threads(args=args)
    if not args.platform.startswith('ilmn'):
        args = check_clair3_options(args)

    args = check_contigs_intersection(args=args, fai_fn=fai_fn)

    return args


def print_args(args):

    logging("")
    logging("[INFO] CALLER VERSION: {}".format(param.version))
    logging("[INFO] NORMAL BAM FILE PATH: {}".format(args.normal_bam_fn))
    logging("[INFO] TUMOR BAM FILE PATH: {}".format(args.tumor_bam_fn))
    logging("[INFO] REFERENCE FILE PATH: {}".format(args.ref_fn))
    logging("[INFO] PLATFORM: {}".format(args.platform))
    logging("[INFO] THREADS: {}".format(args.threads))
    logging("[INFO] OUTPUT FOLDER: {}".format(args.output_dir))
    logging("[INFO] OUTPUT VCF PATH: {}".format(os.path.join(args.output_dir, args.output_prefix + '.vcf.gz')))
    logging("[INFO] PILEUP MODEL PATH: {}".format(args.pileup_model_path))
    logging("[INFO] FULL-ALIGNMENT MODEL PATH: {}".format(args.full_alignment_model_path))
    logging("[INFO] BED FILE PATH: {}".format(args.bed_fn))
    logging("[INFO] GENOTYPING MODE VCF FILE PATH: {}".format(args.genotyping_mode_vcf_fn))
    logging("[INFO] HYBRID MODE VCF FILE PATH: {}".format(args.hybrid_mode_vcf_fn))
    logging("[INFO] REGION FOR CALLING: {}".format(args.region))
    logging("[INFO] CONTIGS FOR CALLING: {}".format(args.ctg_name))
    logging("[INFO] CONDA BINARY PREFIX: {}".format(args.conda_prefix))
    logging("[INFO] SAMTOOLS BINARY PATH: {}".format(args.samtools))
    logging("[INFO] PYTHON BINARY PATH: {}".format(args.python))
    logging("[INFO] PYPY BINARY PATH: {}".format(args.pypy))
    logging("[INFO] PARALLEL BINARY PATH: {}".format(args.parallel))
    logging("[INFO] LONGPHASE BINARY PATH: {}".format(args.longphase))
    logging("[INFO] CHUNK SIZE: {}".format(args.chunk_size))
    logging("[INFO] SNV MINIMUM AF: {}".format(args.snv_min_af))
    logging("[INFO] SNV MINIMUM QUAL: {}".format(args.snv_min_qual))
    if args.enable_indel_calling:
        logging("[INFO] INDEL MINIMUM AF: {}".format(args.indel_min_af))
        logging("[INFO] INDEL PILEUP MODEL PATH: {}".format(args.indel_pileup_model_path))
        logging("[INFO] INDEL FULL-ALIGNMENT MODEL PATH: {}".format(args.indel_full_alignment_model_path))
        logging("[INFO] INDEL MINIMUM QUAL: {}".format(args.indel_min_qual))
    logging("[INFO] NORMAL VCF FILE PATH: {}".format(args.normal_vcf_fn))
    logging("[INFO] DISABLE PHASING: {}".format(args.disable_phasing))
    logging("[INFO] ENABLE DRY RUN: {}".format(args.dry_run))
    logging("[INFO] ENABLE INDEL CALLING: {}".format(args.enable_indel_calling))
    logging("[INFO] ENABLE PRINTING REFERENCE CALLS: {}".format(args.print_ref_calls))
    logging("[INFO] ENABLE PRINTING GERMLINE CALLS: {}".format(args.print_germline_calls))
    logging("[INFO] ENABLE INCLUDING ALL CTGS FOR CALLING: {}".format(args.include_all_ctgs))
    logging("[INFO] ENABLE REMOVING INTERMEDIATE FILES: {}".format(args.remove_intermediate_dir))
    logging("")

    if args.platform.startswith('ilmn'):
        args.platform = 'ilmn'

    if args.platform.startswith('ont'):
        args.platform = 'ont'

    if args.platform.startswith('hifi'):
        args.platform = 'hifi'

    if args.cmdline is not None and args.cmdline != "":
        with open(args.output_dir + '/tmp/CMD', 'w') as f:
            f.write(args.cmdline + '\n')

    return args

def print_command_line(args):

    try:

        cmdline = os.path.realpath(__file__)
        cmdline += ' --tumor_bam_fn {} '.format(args.tumor_bam_fn)
        cmdline += '--normal_bam_fn {} '.format(args.normal_bam_fn)
        cmdline += '--ref_fn {} '.format(args.ref_fn)
        cmdline += '--threads {} '.format(args.threads)
        cmdline += '--platform {} '.format(args.platform)
        cmdline += '--output_dir {} '.format(args.output_dir)
        cmdline += '--pileup_model_path {} '.format(args.pileup_model_path) if args.pileup_model_path is not None else ""
        cmdline += '--full_alignment_model_path {} '.format(args.full_alignment_model_path) if args.full_alignment_model_path is not None else ""
        cmdline += '--ctg_name {} '.format(args.ctg_name) if args.ctg_name is not None else ""
        cmdline += '--region {} '.format(args.region) if args.region is not None else ""
        cmdline += '--bed_fn {} '.format(args.bed_fn) if args.bed_fn is not None else ""
        cmdline += '--genotyping_mode_vcf_fn {} '.format(args.genotyping_mode_vcf_fn) if args.genotyping_mode_vcf_fn is not None else ""
        cmdline += '--hybrid_mode_vcf_fn {} '.format(args.hybrid_mode_vcf_fn) if args.hybrid_mode_vcf_fn is not None else ""
        cmdline += '--qual {} '.format(args.qual) if args.qual is not None else ""
        cmdline += '--snv_min_qual {} '.format(args.snv_min_qual) if args.snv_min_qual is not None else ""
        cmdline += '--indel_min_qual {} '.format(args.indel_min_qual) if args.indel_min_qual is not None else ""
        cmdline += '--snv_min_af {} '.format(args.snv_min_af) if args.snv_min_af is not None else ""
        cmdline += '--indel_min_af {} '.format(args.indel_min_af) if args.indel_min_af is not None else ""
        cmdline += '--indel_max_af_in_normal {} '.format(args.indel_max_af_in_normal) if args.indel_max_af_in_normal is not None else ""
        cmdline += '--min_coverage {} '.format(args.min_coverage) if args.min_coverage is not None else ""
        cmdline += '--chunk_size {} '.format(args.chunk_size) if args.chunk_size is not None else ""
        cmdline += '--sample_name {} '.format(args.sample_name) if args.sample_name != "SAMPLE" else ""
        cmdline += '--output_prefix {} '.format(args.output_prefix) if args.output_prefix != "output" else ""
        cmdline += '--remove_intermediate_dir ' if args.remove_intermediate_dir else ""
        cmdline += '--include_all_ctgs ' if args.include_all_ctgs else ""
        cmdline += '--print_ref_calls ' if args.print_ref_calls else ""
        cmdline += '--print_germline_calls ' if args.print_germline_calls else ""
        cmdline += '--python {} '.format(args.python) if args.python != "python3" else ""
        cmdline += '--pypy {} '.format(args.pypy) if args.pypy != "pypy3" else ""
        cmdline += '--samtools {} '.format(args.samtools) if args.samtools != "samtools" else ""
        cmdline += '--parallel {} '.format(args.parallel) if args.parallel != "parallel" else ""
        cmdline += '--normal_vcf_fn {} '.format(args.normal_vcf_fn) if args.normal_vcf_fn is not None else ""
        cmdline += '--disable_phasing ' if args.disable_phasing else ""
        cmdline += '--haplotagged_tumor_bam_provided_so_skip_intermediate_phasing_and_haplotagging ' if args.haplotagged_tumor_bam_provided_so_skip_intermediate_phasing_and_haplotagging else ""
        cmdline += '--enable_indel_calling ' if args.enable_indel_calling else ""
        cmdline += '--indel_output_prefix {} '.format(args.indel_output_prefix) if args.indel_output_prefix != "indel" else ""
        cmdline += '--indel_pileup_model_path {} '.format(args.indel_pileup_model_path) if args.indel_pileup_model_path is not None else ""
        cmdline += '--indel_full_alignment_model_path {} '.format(args.indel_full_alignment_model_path) if args.indel_full_alignment_model_path is not None else ""
        cmdline += '--clair3_path {} '.format(args.clair3_path) if args.clair3_path is not None else ""
        cmdline += '--clair3_model_path {} '.format(args.clair3_model_path) if args.clair3_model_path is not None else ""
        cmdline += '--longphase {} '.format(args.longphase) if args.longphase is not None else ""
        cmdline += '--whatshap {} '.format(args.whatshap) if args.whatshap is not None else ""
        cmdline += '--whatshap_for_phasing {} '.format(args.whatshap_for_phasing) if args.whatshap_for_phasing is not None else ""
        cmdline += '--prefer_recall ' if args.prefer_recall else ""
        cmdline += '--prefer_balance ' if args.prefer_balance else ""
        cmdline += '--phase_normal {} '.format(args.phase_normal) if args.phase_normal is True else ""
        cmdline += '--phase_tumor {} '.format(args.phase_tumor) if args.phase_tumor is not None else ""
        cmdline += '--use_gpu ' if args.use_gpu else ""
        cmdline += '--indel_min_af {} '.format(args.indel_min_af) if args.indel_min_af is not None else ""
        cmdline += '--enable_realignment False ' if args.enable_realignment is False else ""
        cmdline += '--apply_post_processing False ' if args.apply_post_processing is False else ""
        cmdline += '--skip_steps {} '.format(args.skip_steps) if args.skip_steps is not None else ""
        cmdline += '--clair3_min_coverage {} '.format(args.clair3_min_coverage) if args.clair3_min_coverage is not None else ""
        cmdline += '--clair3_snp_min_af {} '.format(args.clair3_snp_min_af) if args.clair3_snp_min_af is not None else ""
        cmdline += '--clair3_indel_min_af {} '.format(args.clair3_indel_min_af) if args.clair3_indel_min_af is not None else ""
        cmdline += '--enable_clair3_germline_output ' if args.enable_clair3_germline_output else ""
        cmdline += '--use_heterozygous_snp_in_normal_sample_and_normal_bam_for_intermediate_phasing {} '.format(args.use_heterozygous_snp_in_normal_sample_and_normal_bam_for_intermediate_phasing) if args.use_heterozygous_snp_in_normal_sample_and_normal_bam_for_intermediate_phasing is not None else ""
        cmdline += '--use_heterozygous_snp_in_tumor_sample_and_tumor_bam_for_intermediate_phasing {} '.format(args.use_heterozygous_snp_in_tumor_sample_and_tumor_bam_for_intermediate_phasing) if args.use_heterozygous_snp_in_tumor_sample_and_tumor_bam_for_intermediate_phasing is not None else ""
        cmdline += '--use_heterozygous_snp_in_normal_sample_and_tumor_bam_for_intermediate_phasing {} '.format(args.use_heterozygous_snp_in_normal_sample_and_tumor_bam_for_intermediate_phasing) if args.use_heterozygous_snp_in_normal_sample_and_tumor_bam_for_intermediate_phasing is not None else ""
        cmdline += '--use_heterozygous_snp_in_tumor_sample_and_normal_bam_for_intermediate_phasing {} '.format(args.use_heterozygous_snp_in_tumor_sample_and_normal_bam_for_intermediate_phasing) if args.use_heterozygous_snp_in_tumor_sample_and_normal_bam_for_intermediate_phasing is not None else ""
        cmdline += '--use_heterozygous_snp_in_normal_sample_for_intermediate_phasing {} '.format(args.use_heterozygous_snp_in_normal_sample_for_intermediate_phasing) if args.use_heterozygous_snp_in_normal_sample_for_intermediate_phasing is not None else ""
        cmdline += '--use_heterozygous_snp_in_tumor_sample_for_intermediate_phasing {} '.format(args.use_heterozygous_snp_in_tumor_sample_for_intermediate_phasing) if args.use_heterozygous_snp_in_tumor_sample_for_intermediate_phasing is not None else ""
        cmdline += '--use_heterozygous_indel_for_intermediate_phasing {} '.format(args.use_heterozygous_indel_for_intermediate_phasing) if args.use_heterozygous_indel_for_intermediate_phasing is not None else ""
        cmdline += '--use_normal_bam_for_intermediate_phasing {} '.format(args.use_normal_bam_for_intermediate_phasing) if args.use_normal_bam_for_intermediate_phasing is not None else ""
        cmdline += '--use_longphase_for_intermediate_haplotagging {} '.format(args.use_longphase_for_intermediate_haplotagging) if args.use_longphase_for_intermediate_haplotagging is not None else ""
        cmdline += '--enable_verdict ' if args.enable_verdict else ""
        cmdline += '--conda_prefix {} '.format(args.conda_prefix) if args.conda_prefix is not None else ""
        args.cmdline = cmdline
    except:
        return args
    logging("[COMMAND] " + cmdline + '\n')
    return args

def somatic_calling(args):

    step = 1
    echo_list = []
    commands_list = []
    tmp_vcf_output_path = args.output_path.tmp_vcf_output_path
    vcf_output_path = args.output_path.vcf_output_path
    clair3_output_path = args.output_dir + '/tmp/clair3_output'
    normal_bam_fn = clair3_output_path + '/phased_output/normal_{1/.}.bam' if args.phase_normal else args.normal_bam_fn
    tumor_bam_fn = clair3_output_path + '/phased_output/tumor_{1/.}.bam' if args.phase_tumor else args.tumor_bam_fn
    tumor_bam_prefix = clair3_output_path + '/phased_output/tumor_' if args.phase_tumor else args.tumor_bam_fn

    try:
        rc = subprocess.check_call('time', shell=True, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
        time = 'time '
    except subprocess.CalledProcessError as e:
        time = ''

    if args.clair3_path is not None and args.platform != 'ilmn' and args.phase_tumor:

        if args.normal_vcf_fn:
            normal_vcf_fn = args.normal_vcf_fn
        else:
            normal_vcf_fn = clair3_output_path + '/clair3_normal_output/merge_output.vcf.gz'
            echo_list.append("[INFO] Call Germline Variants in Normal BAM using Clair3")
            clair3_normal_command = '( ' + time + args.clair3_path + '/run_clair3.sh'
            clair3_normal_command += ' --bam_fn ' + args.normal_bam_fn
            clair3_normal_command += ' --ref_fn ' + args.ref_fn
            clair3_normal_command += ' --model_path ' + args.clair3_model_path
            clair3_normal_command += ' --platform ' + args.platform
            clair3_normal_command += ' --threads ' + str(args.threads)
            clair3_normal_command += ' --output ' + clair3_output_path + '/clair3_normal_output'
            clair3_normal_command += ' --ctg_name=' + args.clair3_option.ctg_name_str
            clair3_normal_command += ' --samtools=' + args.samtools
            clair3_normal_command += ' --pypy=' + args.pypy
            clair3_normal_command += ' --python=' + args.python
            clair3_normal_command += ' --min_coverage=' + args.clair3_option.min_coverage
            clair3_normal_command += ' --snp_min_af=' + args.clair3_option.snp_min_af
            clair3_normal_command += ' --indel_min_af=' + args.clair3_option.indel_min_af
            if args.clair3_option.longphase_for_phasing is True:
                clair3_normal_command += ' --longphase_for_phasing '
            clair3_normal_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/clair3_log/1_CLAIR3_NORMAL.log'
            commands_list.append(clair3_normal_command)

        echo_list.append("[INFO] Call Germline Variant in Tumor BAM using Clair3")
        clair3_tumor_command = '( ' + time + args.clair3_path + '/run_clair3.sh'
        clair3_tumor_command += ' --bam_fn ' + args.tumor_bam_fn
        clair3_tumor_command += ' --ref_fn ' + args.ref_fn
        clair3_tumor_command += ' --model_path ' + args.clair3_model_path
        clair3_tumor_command += ' --platform ' + args.platform
        clair3_tumor_command += ' --threads ' + str(args.threads)
        clair3_tumor_command += ' --output ' + clair3_output_path + '/clair3_tumor_output'
        clair3_tumor_command += ' --ctg_name=' + args.clair3_option.ctg_name_str
        clair3_tumor_command += ' --samtools=' + args.samtools
        clair3_tumor_command += ' --pypy=' + args.pypy
        clair3_tumor_command += ' --python=' + args.python
        clair3_tumor_command += ' --min_coverage=' + args.clair3_option.min_coverage
        clair3_tumor_command += ' --snp_min_af=' + args.clair3_option.snp_min_af
        clair3_tumor_command += ' --indel_min_af=' + args.clair3_option.indel_min_af
        if args.clair3_option.longphase_for_phasing is True:
            clair3_tumor_command += ' --longphase_for_phasing '
        clair3_tumor_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/clair3_log/2_CLAIR3_TUMOR.log'
        commands_list.append(clair3_tumor_command)

        echo_list.append("[INFO] Select Heterozygous SNP for Phasing")
        ssp_command = '( ' + time + args.parallel
        ssp_command += ' --joblog ' + args.output_dir + '/logs/clair3_log/parallel_1_select_hetero_snp_for_phasing.log'
        ssp_command += ' -j ' + str(args.threads)
        ssp_command += ' ' + args.pypy + ' ' + main_entry + ' select_hetero_snp_for_phasing'
        ssp_command += ' --tumor_vcf_fn ' + clair3_output_path + '/clair3_tumor_output/merge_output.vcf.gz'
        ssp_command += ' --normal_vcf_fn ' + normal_vcf_fn
        ssp_command += ' --output_folder ' + clair3_output_path + '/vcf'
        ssp_command += ' --ctg_name {1}'
        ssp_command += ' --use_heterozygous_snp_in_normal_sample_for_intermediate_phasing True' if args.use_heterozygous_snp_in_normal_sample_for_intermediate_phasing else ""
        ssp_command += ' --use_heterozygous_snp_in_tumor_sample_for_intermediate_phasing True' if args.use_heterozygous_snp_in_tumor_sample_for_intermediate_phasing else ""
        ssp_command += ' --use_heterozygous_indel_for_intermediate_phasing True' if args.use_heterozygous_indel_for_intermediate_phasing else ""
        ssp_command += ' :::: ' + args.output_dir + '/tmp/CONTIGS'
        ssp_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/clair3_log/1_select_hetero_snp_for_phasing.log'
        commands_list.append(ssp_command)

        if args.phase_normal:
            echo_list.append("[INFO] Phase the Normal BAM")
            if args.clair3_option.longphase_for_phasing is not None:
                pn_command = '( ' + time + args.parallel
                pn_command += ' --joblog ' + args.output_dir + '/logs/clair3_log/parallel_2_phase_normal.log'
                pn_command += ' -j ' + str(args.threads)
                pn_command += ' ' + args.longphase + ' phase '
                pn_command += ' -s ' + clair3_output_path + '/vcf/{1}.vcf'
                pn_command += ' -b ' + args.normal_bam_fn
                pn_command += ' -r ' + args.ref_fn
                pn_command += ' -t ' + str(args.threads)
                pn_command += ' --indels ' if args.use_heterozygous_indel_for_intermediate_phasing else ""
                pn_command += ' -o ' + clair3_output_path + '/phased_output/normal_phased_{1}'
                pn_command += ' --ont' if args.platform == 'ont' else ' --pb'
                pn_command += ' :::: ' + args.output_dir + '/tmp/CONTIGS'
                pn_command += ' && ' + args.parallel
                pn_command += ' -j ' + str(args.threads)
                pn_command += ' bgzip -f ' + clair3_output_path + '/phased_output/normal_phased_{1}.vcf'
                pn_command += ' :::: ' + args.output_dir + '/tmp/CONTIGS'
                pn_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/clair3_log/2_phase_normal.log'
            else:
                pn_command = '( ' + time + args.parallel
                pn_command += ' --joblog ' + args.output_dir + '/logs/clair3_log/parallel_2_phase_normal.log'
                pn_command += ' -j ' + str(args.threads)
                pn_command += ' ' + args.whatshap + ' phase '
                pn_command += ' --output ' + clair3_output_path + '/phased_output/normal_phased_{1}.vcf.gz'
                pn_command += ' --reference ' + args.ref_fn
                pn_command += ' --chromosome {1}'
                pn_command += ' --distrust-genotypes'
                pn_command += ' --ignore-read-groups'
                pn_command += ' ' + clair3_output_path + '/vcf/{1}.vcf'
                pn_command += ' :::: ' + args.output_dir + '/tmp/CONTIGS'
                pn_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/clair3_log/2_phase_normal.log'

            tabix_command = args.parallel + ' -j ' + str(args.threads)
            tabix_command += ' tabix' + ' -f -p vcf'
            tabix_command += ' ' + clair3_output_path + '/phased_output/normal_phased_{1}.vcf.gz'
            tabix_command += ' :::: ' + args.output_dir + '/tmp/CONTIGS'
            commands_list.append(pn_command + ' && ' + tabix_command)

            echo_list.append("[INFO] Haplotag the Normal BAM")
            ht_command = '( ' + time + args.parallel
            ht_command += ' --joblog ' + args.output_dir + '/logs/parallel_3_haplotag_normal.log'
            ht_command += ' -j ' + str(args.threads)
            ht_command += ' ' + args.whatshap + ' haplotag'
            ht_command += ' --output ' + clair3_output_path + '/phased_output/normal_{1}.bam'
            ht_command += ' --reference ' + args.ref_fn
            ht_command += ' --regions {1} '
            ht_command += ' --ignore-read-groups'
            ht_command += ' ' + clair3_output_path + '/phased_output/normal_phased_{1}.vcf.gz'
            ht_command += ' ' + args.normal_bam_fn
            ht_command += ' :::: ' + args.output_dir + '/tmp/CONTIGS'
            ht_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/3_normal_haplotag.log'

            index_command = args.parallel + ' -j ' + str(args.threads)
            index_command += ' ' + args.samtools + ' index '
            index_command += ' -@' + str(args.threads)
            index_command += ' ' + clair3_output_path + '/phased_output/normal_{1}.bam'
            index_command += ' :::: ' + args.output_dir + '/tmp/CONTIGS'
            commands_list.append(ht_command + ' && ' + index_command)

        echo_list.append("[INFO] Phase the Tumor BAM")
        if args.clair3_option.longphase_for_phasing is not None:
            pt_command = '( ' + time + args.parallel
            pt_command += ' --joblog ' + args.output_dir + '/logs/clair3_log/parallel_4_phase_tumor.log'
            pt_command += ' -j ' + str(args.threads)
            pt_command += ' ' + args.clair3_option.longphase + ' phase '
            pt_command += ' -s ' + clair3_output_path + '/vcf/{1}.vcf'
            pt_command += ' -b ' + (args.tumor_bam_fn if not args.use_normal_bam_for_intermediate_phasing else args.normal_bam_fn)
            pt_command += ' -r ' + args.ref_fn
            pt_command += ' -t ' + str(args.threads)
            pt_command += ' --indels ' if args.use_heterozygous_indel_for_intermediate_phasing else ""
            pt_command += ' -o ' + clair3_output_path + '/phased_output/tumor_phased_{1}'
            pt_command += ' --ont' if args.platform == 'ont' else ' --pb'
            pt_command += ' :::: ' + args.output_dir + '/tmp/CONTIGS'
            pt_command += ' && ' + args.parallel
            pt_command += ' -j ' + str(args.threads)
            pt_command += ' bgzip -f ' + clair3_output_path + '/phased_output/tumor_phased_{1}.vcf'
            pt_command += ' :::: ' + args.output_dir + '/tmp/CONTIGS'
            pt_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/clair3_log/4_phase_tumor.log'
        else:
            pt_command = '( ' + time + args.parallel
            pt_command += ' --joblog ' + args.output_dir + '/logs/clair3_log/parallel_4_phase_tumor.log'
            pt_command += ' -j ' + str(args.threads)
            pt_command += ' ' + args.whatshap + ' phase '
            pt_command += ' --output ' + clair3_output_path + '/phased_output/tumor_phased_{1}.vcf.gz'
            pt_command += ' --reference ' + args.ref_fn
            pt_command += ' --chromosome {1}'
            pt_command += ' --distrust-genotypes'
            pt_command += ' --ignore-read-groups'
            pt_command += ' ' + clair3_output_path + '/vcf/{1}.vcf'
            pt_command += ' :::: ' + args.output_dir + '/tmp/CONTIGS'
            pt_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/clair3_log/4_phase_tumor.log'

        tabix_command = args.parallel + ' -j ' + str(args.threads)
        tabix_command += ' tabix' + ' -f -p vcf'
        tabix_command += ' ' + clair3_output_path + '/phased_output/tumor_phased_{1}.vcf.gz'
        tabix_command += ' :::: ' + args.output_dir + '/tmp/CONTIGS'
        commands_list.append(pt_command + ' && ' + tabix_command)

        echo_list.append("[INFO] Haplotag the Tumor BAM")
        if args.use_longphase_for_intermediate_haplotagging:
            ht_command = '( ' + time + args.parallel
            ht_command += ' --joblog ' + args.output_dir + '/logs/clair3_log/parallel_5_haplotag_tumor.log'
            ht_command += ' -j ' + str(args.threads)
            ht_command += ' ' + args.longphase + ' haplotag'
            ht_command += ' -o ' + clair3_output_path + '/phased_output/tumor_{1}'
            ht_command += ' --reference ' + args.ref_fn
            ht_command += ' --region {1} '
            ht_command += ' -s ' + clair3_output_path + '/phased_output/tumor_phased_{1}.vcf.gz'
            ht_command += ' -b ' + args.tumor_bam_fn
            ht_command += ' :::: ' + args.output_dir + '/tmp/CONTIGS'
            ht_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/clair3_log/5_tumor_haplotag.log'
        else:
            ht_command = '( ' + time + args.parallel
            ht_command += ' --joblog ' + args.output_dir + '/logs/clair3_log/parallel_5_haplotag_tumor.log'
            ht_command += ' -j ' + str(args.threads)
            ht_command += ' ' + args.whatshap + ' haplotag'
            ht_command += ' --output ' + clair3_output_path + '/phased_output/tumor_{1}.bam'
            ht_command += ' --reference ' + args.ref_fn
            ht_command += ' --regions {1} '
            ht_command += ' --ignore-read-groups'
            ht_command += ' ' + clair3_output_path + '/phased_output/tumor_phased_{1}.vcf.gz'
            ht_command += ' ' + args.tumor_bam_fn
            ht_command += ' :::: ' + args.output_dir + '/tmp/CONTIGS'
            ht_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/clair3_log/5_tumor_haplotag.log'

        index_command = args.parallel + ' -j ' + str(args.threads)
        index_command += ' ' + args.samtools + ' index '
        index_command += ' -@' + str(args.threads)
        index_command += ' ' + clair3_output_path + '/phased_output/tumor_{1}.bam'
        index_command += ' :::: ' + args.output_dir + '/tmp/CONTIGS'
        commands_list.append(ht_command + ' && ' + index_command)

    # Pileup calling
    #STEP 1: EXTRACT CANDIDATES
    echo_list.append("[INFO] STEP {}: Extract Variant Candidates from Tumor and Normal BAMs".format(step))
    step += 1
    ec_command = '( ' + time + args.parallel
    ec_command += ' --joblog ' + args.output_dir + '/logs/parallel_1_extract_tumor_candidates.log'
    ec_command += ' -C " " -j ' + str(args.threads)
    ec_command += ' ' + args.pypy + ' ' + main_entry + ' extract_pair_candidates'
    ec_command += ' --tumor_bam_fn ' + args.tumor_bam_fn
    ec_command += ' --normal_bam_fn ' + args.normal_bam_fn
    ec_command += ' --ref_fn ' + args.ref_fn
    ec_command += ' --samtools ' + args.samtools
    ec_command += ' --snv_min_af ' + str(args.snv_min_af)
    ec_command += ' --indel_min_af ' + str(args.indel_min_af)
    ec_command += ' --chunk_id {2} '
    ec_command += ' --chunk_num {3} '
    ec_command += ' --ctg_name {1} '
    ec_command += ' --platform ' + args.platform
    ec_command += ' --min_coverage ' + str(args.min_coverage)
    ec_command += (' --min_bq ' + str(args.min_bq)) if args.min_bq is not None else ""
    ec_command += ' --bed_fn ' + os.path.join(args.output_dir, 'tmp', 'split_beds', '{1}')
    ec_command += ' --candidates_folder ' + args.output_dir + '/tmp/candidates'
    ec_command += ' --output_depth True '
    ec_command += ' --select_indel_candidates ' + str(args.enable_indel_calling)
    ec_command += ' --hybrid_mode_vcf_fn ' + str(args.hybrid_mode_vcf_fn)
    ec_command += ' --genotyping_mode_vcf_fn ' + str(args.genotyping_mode_vcf_fn)
    ec_command += ' --enable_params_for_liquid_tumor_sample True' if args.enable_params_for_liquid_tumor_sample else ""
    ec_command += ' :::: ' + os.path.join(args.output_dir, 'tmp', 'CHUNK_LIST')
    ec_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/1_EC.log'
    ec_command += ' && ' + args.pypy + ' ' + main_entry + ' concat_files'
    ec_command += ' --input_dir ' + "{}/tmp/candidates".format(args.output_dir)
    ec_command += ' --input_prefix ' + "CANDIDATES_FILE_"
    ec_command += ' --output_fn CANDIDATES_FILES '
    commands_list.append(ec_command)

    ##STEP 2: CREATE PAIR TENSOR
    echo_list.append("[INFO] STEP 2: Pileup Model Calling\n")
    echo_list[-1] += ("[INFO] Create Paired Tensors")
    cpt_command = '( ' + time + args.parallel
    cpt_command += ' --joblog ' + args.output_dir + '/logs/parallel_2-1_create_pair_tensor.log'
    cpt_command += ' -j ' + str(args.threads)
    cpt_command += ' ' + args.pypy + ' ' + main_entry + ' create_pair_tensor_pileup'
    cpt_command += ' --normal_bam_fn ' + args.normal_bam_fn
    cpt_command += ' --tumor_bam_fn ' + args.tumor_bam_fn
    cpt_command += ' --ref_fn ' + args.ref_fn
    cpt_command += ' --ctg_name {1/.}'
    cpt_command += (' --min_bq ' + str(args.min_bq)) if args.min_bq is not None else ""
    cpt_command += ' --samtools ' + args.samtools
    cpt_command += ' --candidates_bed_regions {1}'
    cpt_command += ' --tensor_can_fn ' + args.output_dir + '/tmp/pileup_tensor_can/{1/} '
    cpt_command += ' --platform ' + args.platform
    cpt_command += ' :::: ' + args.output_dir + '/tmp/candidates/CANDIDATES_FILES'
    cpt_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/2-1_CPT.log'
    commands_list += [cpt_command]

    ## STEP 3: PREDICT
    echo_list.append("[INFO] Pileup Model Prediction")
    p_predict_command = '( ' + time + args.parallel
    p_predict_command += ' --joblog ' + args.output_dir + '/logs/parallel_2-2_predict.log'
    p_predict_command += ' -j ' + str(args.threads)
    p_predict_command += ' ' + args.python + ' ' + main_entry + ' predict'
    p_predict_command += ' --tensor_fn ' + args.output_dir + '/tmp/pileup_tensor_can/{1/} '
    p_predict_command += ' --call_fn ' + args.output_dir + '/tmp/vcf_output/p_{1/}.vcf'
    p_predict_command += ' --chkpnt_fn ' + args.pileup_model_path
    p_predict_command += ' --use_gpu ' + str(args.use_gpu)
    p_predict_command += ' --platform ' + args.platform
    p_predict_command += ' --ctg_name {1/.}'
    p_predict_command += ' --pileup '
    p_predict_command += ' --show_ref ' if args.print_ref_calls else ""
    p_predict_command += ' --show_germline ' if args.print_germline_calls else ""
    p_predict_command += ' :::: ' + args.output_dir + '/tmp/candidates/CANDIDATES_FILES'
    p_predict_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/2-2_PREDICT.log'
    commands_list += [p_predict_command]

    # STEP 4: MERGE VCF
    echo_list.append("[INFO] Merge Pileup VCFs")
    p_mv_command = args.pypy + ' ' + main_entry + ' sort_vcf'
    p_mv_command += ' --ref_fn ' + args.ref_fn
    p_mv_command += ' --contigs_fn ' + os.path.join(args.output_dir, 'tmp', 'CONTIGS')
    p_mv_command += ' --input_dir ' + args.output_dir + '/tmp/vcf_output'
    p_mv_command += ' --vcf_fn_prefix ' + 'p_'
    p_mv_command += ' --output_fn ' + args.output_dir + '/tmp/vcf_output/pileup.vcf'
    commands_list += [p_mv_command]

    # ## Full-alignment calling
    normal_bam_fn = clair3_output_path + '/phased_output/normal_{1/.}.bam' if args.phase_normal else args.normal_bam_fn
    tumor_bam_fn = clair3_output_path + '/phased_output/tumor_{1/.}.bam' if args.phase_tumor else args.tumor_bam_fn
    tumor_bam_prefix = clair3_output_path + '/phased_output/tumor_' if args.phase_tumor else args.tumor_bam_fn

    echo_list.append("[INFO] STEP 3: Full-alignment Model Calling\n")
    echo_list[-1] += "[INFO] Create Full-alignment Paired Tensors"
    cpt_fa_command = '( ' + time + args.parallel
    cpt_fa_command += ' --joblog ' + args.output_dir + '/logs/parallel_3-1_create_pair_tensor_fa.log'
    cpt_fa_command += ' -j ' + str(args.threads)
    cpt_fa_command += ' ' + args.pypy + ' ' + main_entry + ' create_pair_tensor'
    cpt_fa_command += ' --normal_bam_fn ' + normal_bam_fn
    cpt_fa_command += ' --tumor_bam_fn ' + tumor_bam_fn
    cpt_fa_command += ' --ref_fn ' + args.ref_fn
    cpt_fa_command += ' --ctg_name {1/.}'
    cpt_fa_command += ' --samtools ' + args.samtools
    cpt_fa_command += ' --candidates_bed_regions {1}'
    cpt_fa_command += ' --tensor_can_fn ' + args.output_dir + '/tmp/fa_tensor_can/{1/} '
    cpt_fa_command += ' --platform ' + args.platform
    cpt_fa_command += ' :::: ' + args.output_dir + '/tmp/candidates/CANDIDATES_FILES'
    cpt_fa_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/3-1_CPT.log'
    commands_list += [cpt_fa_command]

    ## STEP 3: PREDICT
    echo_list.append("[INFO] Full-alignment Model Prediction")
    fa_predict_command = '( ' + time + args.parallel
    fa_predict_command += ' --joblog ' + args.output_dir + '/logs/parallel_3-2_predict.log'
    fa_predict_command += ' -j ' + str(args.threads)
    fa_predict_command += ' ' + args.python + ' ' + main_entry + ' predict'
    fa_predict_command += ' --tensor_fn ' + args.output_dir + '/tmp/fa_tensor_can/{1/} '
    fa_predict_command += ' --call_fn ' + args.output_dir + '/tmp/vcf_output/fa_{1/}.vcf'
    fa_predict_command += ' --chkpnt_fn ' + args.full_alignment_model_path
    fa_predict_command += ' --use_gpu ' + str(args.use_gpu)
    fa_predict_command += ' --platform ' + args.platform
    fa_predict_command += ' --ctg_name {1/.}'
    fa_predict_command += ' --show_ref ' if args.print_ref_calls else ""
    fa_predict_command += ' --show_germline ' if args.print_germline_calls else ""
    fa_predict_command += ' :::: ' + args.output_dir + '/tmp/candidates/CANDIDATES_FILES'
    fa_predict_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/3-2_PREDICT.log'
    commands_list += [fa_predict_command]

    ## STEP 4: MERGE VCF
    echo_list.append("[INFO] Merge Full-alignment VCFs")
    fa_mv_command = args.pypy + ' ' + main_entry + ' sort_vcf'
    fa_mv_command += ' --ref_fn ' + args.ref_fn
    fa_mv_command += ' --contigs_fn ' + os.path.join(args.output_dir, 'tmp', 'CONTIGS')
    fa_mv_command += ' --input_dir ' + args.output_dir + '/tmp/vcf_output'
    fa_mv_command += ' --vcf_fn_prefix ' + 'fa_'
    fa_mv_command += ' --output_fn ' + args.output_dir + '/tmp/vcf_output/full_alignment.vcf'
    commands_list += [fa_mv_command]

    # short-read realignment
    if args.platform == 'ilmn':
        echo_list.append("[INFO] STEP 4: Short-read realignment")
        realign_command = '( ' + time + args.python + ' ' + main_entry + ' realign_variants'
        realign_command += ' --bam_fn ' + args.tumor_bam_fn
        realign_command += ' --ref_fn ' + args.ref_fn
        realign_command += ' --pileup_vcf_fn ' + args.output_dir + '/tmp/vcf_output/pileup.vcf'
        realign_command += ' --full_alignment_vcf_fn ' + args.output_dir + '/tmp/vcf_output/full_alignment.vcf'
        realign_command += ' --output_dir ' + args.output_dir + '/tmp/vcf_output'
        realign_command += ' --samtools ' + args.samtools
        realign_command += ' --python ' + args.python
        realign_command += ' --threads ' + str(args.threads)
        realign_command += ' --enable_realignment ' + str(args.enable_realignment)
        realign_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/4_REALIGN.log'
        commands_list += [realign_command]

    #graph postprocessing
    else:
        echo_list.append("[INFO] STEP 4: Haplotype filtering")
        hap_g_command = '( ' + time + args.pypy + ' ' + main_entry + ' haplotype_filtering'
        hap_g_command += ' --tumor_bam_fn ' + tumor_bam_prefix
        hap_g_command += ' --ref_fn ' + args.ref_fn
        hap_g_command += ' --germline_vcf_fn ' + clair3_output_path + '/clair3_tumor_output/merge_output.vcf.gz'
        hap_g_command += ' --pileup_vcf_fn ' + args.output_dir + '/tmp/vcf_output/pileup.vcf'
        hap_g_command += ' --full_alignment_vcf_fn ' + args.output_dir + '/tmp/vcf_output/full_alignment.vcf'
        hap_g_command += ' --output_dir ' + args.output_dir + '/tmp/vcf_output'
        hap_g_command += ' --samtools ' + args.samtools
        hap_g_command += ' --pypy3 ' + args.pypy
        hap_g_command += ' --parallel ' + args.parallel
        hap_g_command += ' --threads ' + str(args.threads)
        hap_g_command += ' --debug ' if args.debug else ''
        hap_g_command += ' --show_ref ' if args.print_ref_calls else ''
        hap_g_command += ' --apply_post_processing False' if not args.apply_post_processing else ''
        hap_g_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/4_HAP_FILTER.log'
        commands_list += [hap_g_command]

    echo_list.append("[INFO] STEP 5: Merge and sort VCF")
    sort_vcf_command = '( ' + time + args.pypy + ' ' + main_entry + ' merge_vcf'
    sort_vcf_command += ' --ref_fn ' + args.ref_fn
    sort_vcf_command += ' --pileup_vcf_fn ' + args.output_dir + '/tmp/vcf_output/pileup_filter.vcf'
    sort_vcf_command += ' --full_alignment_vcf_fn ' + args.output_dir + '/tmp/vcf_output/full_alignment_filter.vcf'
    sort_vcf_command += ' --output_fn ' + args.output_dir + '/{}.vcf'.format(args.output_prefix)
    sort_vcf_command += ' --platform ' + args.platform
    sort_vcf_command += ' --qual ' + str(args.snv_min_qual)
    sort_vcf_command += ' --sample_name ' + str(args.sample_name)
    sort_vcf_command += ' --enable_indel_calling ' + str(args.enable_indel_calling)
    sort_vcf_command += ' --prefer_recall ' + str(args.prefer_recall)
    sort_vcf_command += ' --cmdline ' + args.output_dir + '/tmp/CMD'
    sort_vcf_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/5_MV.log'
    commands_list += [sort_vcf_command]

    if args.genotyping_mode_vcf_fn is not None or args.hybrid_mode_vcf_fn is not None:
        echo_list.append("[INFO] Add reference calls to the output VCF output")
        genotyping_command = args.pypy + ' ' + main_entry + ' add_back_missing_variants_in_genotyping'
        genotyping_command += ' --genotyping_mode_vcf_fn ' + str(args.genotyping_mode_vcf_fn)
        genotyping_command += ' --hybrid_mode_vcf_fn ' + str(args.hybrid_mode_vcf_fn)
        genotyping_command += ' --call_fn ' + args.output_dir + '/{}.vcf.gz'.format(args.output_prefix)
        genotyping_command += ' --output_fn ' + args.output_dir + '/{}.vcf'.format(args.output_prefix)
        genotyping_command += ' --candidates_folder ' + args.output_dir + '/tmp/candidates'
        genotyping_command += ' 2>&1 | tee ' + args.output_dir + '/logs/6_GT.log'
        commands_list += [genotyping_command]

    if args.enable_verdict:
        echo_list.append("[INFO] Add CNV germline tagging to output VCF")
        cnv_germline_tagging_command = args.python + ' ' + main_entry + ' cnv_germline_tagging'
        cnv_germline_tagging_command += ' --tumor_bam_fn ' + args.tumor_bam_fn
        cnv_germline_tagging_command += ' --normal_bam_fn ' + args.normal_bam_fn
        cnv_germline_tagging_command += ' --input_vcf_fn ' + args.output_dir + '/{}.vcf.gz'.format(args.output_prefix)
        cnv_germline_tagging_command += ' --allele_counter ' + str(args.allele_counter_dir)
        cnv_germline_tagging_command += ' --cnv_resource_dir ' + str(args.cnv_resource_dir)
        cnv_germline_tagging_command += ' --output_fn ' + args.output_dir + '/{}_cnv_germline_tagged.vcf'.format(args.output_prefix)
        cnv_germline_tagging_command += ' --output_dir ' + args.output_dir + '/tmp/cnv_output'
        cnv_germline_tagging_command += ' --parallel ' + args.parallel
        cnv_germline_tagging_command += ' --python ' + args.python
        cnv_germline_tagging_command += ' --contig_fn ' + args.output_dir + '/tmp/CONTIGS'
        cnv_germline_tagging_command += ' --threads ' + str(args.threads)
        cnv_germline_tagging_command += ' 2>&1 | tee ' + args.output_dir + '/logs/7_CGT.log'
        commands_list += [cnv_germline_tagging_command]


    if args.enable_indel_calling:
        ##STEP 2: CREATE PAIR TENSOR
        echo_list.append("[INFO] STEP 6: Indel Pileup Model Calling\n")
        echo_list[-1] += ("[INFO] Create Paired Tensors")
        indel_cpt_command = args.pypy + ' ' + main_entry + ' concat_files'
        indel_cpt_command += ' --input_dir ' + "{}/tmp/candidates".format(args.output_dir)
        indel_cpt_command += ' --input_prefix ' + "INDEL_CANDIDATES_FILE_"
        indel_cpt_command += ' --output_fn INDEL_CANDIDATES_FILES '
        indel_cpt_command += ' && ( ' + time + args.parallel
        indel_cpt_command += ' --joblog ' + args.output_dir + '/logs/parallel_6-1_create_pair_tensor_indel.log'
        indel_cpt_command += ' -j ' + str(args.threads)
        indel_cpt_command += ' ' + args.pypy + ' ' + main_entry + ' create_pair_tensor_pileup'
        indel_cpt_command += ' --normal_bam_fn ' + args.normal_bam_fn
        indel_cpt_command += ' --tumor_bam_fn ' + args.tumor_bam_fn
        indel_cpt_command += ' --ref_fn ' + args.ref_fn
        indel_cpt_command += ' --ctg_name {1/.}'
        indel_cpt_command += ' --samtools ' + args.samtools
        indel_cpt_command += ' --candidates_bed_regions {1}'
        indel_cpt_command += ' --tensor_can_fn ' + args.output_dir + '/tmp/pileup_tensor_can/indel_{1/} '
        indel_cpt_command += ' --platform ' + args.platform
        indel_cpt_command += ' :::: ' + args.output_dir + '/tmp/candidates/INDEL_CANDIDATES_FILES'
        indel_cpt_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/6-1_CPTI.log'
        commands_list += [indel_cpt_command]

        ## INDEL PREDICT
        echo_list.append("[INFO] Indel Pileup Model Prediction")
        indel_p_predict_command = '( ' + time + args.parallel
        indel_p_predict_command += ' --joblog ' + args.output_dir + '/logs/parallel_6-2_predict_indel.log'
        indel_p_predict_command += ' -j ' + str(args.threads)
        indel_p_predict_command += ' ' + args.python + ' ' + main_entry + ' predict'
        indel_p_predict_command += ' --tensor_fn ' + args.output_dir + '/tmp/pileup_tensor_can/indel_{1/} '
        indel_p_predict_command += ' --call_fn ' + args.output_dir + '/tmp/vcf_output/indel_p_{1/}.vcf'
        indel_p_predict_command += ' --chkpnt_fn ' + args.indel_pileup_model_path
        indel_p_predict_command += ' --use_gpu ' + str(args.use_gpu)
        indel_p_predict_command += ' --platform ' + args.platform
        indel_p_predict_command += ' --ctg_name {1/.}'
        indel_p_predict_command += ' --pileup '
        indel_p_predict_command += ' --enable_indel_calling True '
        indel_p_predict_command += ' --show_ref ' if args.print_ref_calls else ""
        indel_p_predict_command += ' --show_germline ' if args.print_germline_calls else ""
        indel_p_predict_command += ' :::: ' + args.output_dir + '/tmp/candidates/INDEL_CANDIDATES_FILES'
        indel_p_predict_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/6-2_PREDICT_INDEL.log'
        commands_list += [indel_p_predict_command]

        # MERGE INDEL VCF
        echo_list.append("[INFO] Merge Pileup VCFs")
        indel_p_mv_command = args.pypy + ' ' + main_entry + ' sort_vcf'
        indel_p_mv_command += ' --ref_fn ' + args.ref_fn
        indel_p_mv_command += ' --contigs_fn ' + os.path.join(args.output_dir, 'tmp', 'CONTIGS')
        indel_p_mv_command += ' --input_dir ' + args.output_dir + '/tmp/vcf_output'
        indel_p_mv_command += ' --vcf_fn_prefix ' + 'indel_p_'
        indel_p_mv_command += ' --output_fn ' + args.output_dir + '/tmp/vcf_output/indel_pileup.vcf'
        commands_list += [indel_p_mv_command]

        echo_list.append("[INFO] STEP 7: Indel Full-alignment Model Calling\n")
        echo_list[-1] += "[INFO] Create Full-alignment Paired Tensors"
        indel_cpt_fa_command = '( ' + time + args.parallel
        indel_cpt_fa_command += ' --joblog ' + args.output_dir + '/logs/parallel_7-1_create_pair_tensor_fa_indel.log'
        indel_cpt_fa_command += ' -j ' + str(args.threads)
        indel_cpt_fa_command += ' ' + args.pypy + ' ' + main_entry + ' create_pair_tensor'
        indel_cpt_fa_command += ' --normal_bam_fn ' + normal_bam_fn
        indel_cpt_fa_command += ' --tumor_bam_fn ' + tumor_bam_fn
        indel_cpt_fa_command += ' --ref_fn ' + args.ref_fn
        indel_cpt_fa_command += ' --ctg_name {1/.}'
        indel_cpt_fa_command += ' --samtools ' + args.samtools
        indel_cpt_fa_command += ' --candidates_bed_regions {1}'
        indel_cpt_fa_command += ' --tensor_can_fn ' + args.output_dir + '/tmp/fa_tensor_can/indel_{1/} '
        indel_cpt_fa_command += ' --platform ' + args.platform
        indel_cpt_fa_command += ' :::: ' + args.output_dir + '/tmp/candidates/INDEL_CANDIDATES_FILES'
        indel_cpt_fa_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/7-1_CPTI.log'
        commands_list += [indel_cpt_fa_command]

        ## STEP 3: INDEL PREDICT
        echo_list.append("[INFO] Indel Full-alignment Model Prediction")
        indel_fa_predict_command = '( ' + time + args.parallel
        indel_fa_predict_command += ' --joblog ' + args.output_dir + '/logs/parallel_7-2_predict.log'
        indel_fa_predict_command += ' -j ' + str(args.threads)
        indel_fa_predict_command += ' ' + args.python + ' ' + main_entry + ' predict'
        indel_fa_predict_command += ' --tensor_fn ' + args.output_dir + '/tmp/fa_tensor_can/indel_{1/} '
        indel_fa_predict_command += ' --call_fn ' + args.output_dir + '/tmp/vcf_output/indel_fa_{1/}.vcf'
        indel_fa_predict_command += ' --chkpnt_fn ' + args.indel_full_alignment_model_path
        indel_fa_predict_command += ' --use_gpu ' + str(args.use_gpu)
        indel_fa_predict_command += ' --platform ' + args.platform
        indel_fa_predict_command += ' --ctg_name {1/.}'
        indel_fa_predict_command += ' --enable_indel_calling True '
        indel_fa_predict_command += ' --show_ref ' if args.print_ref_calls else ""
        indel_fa_predict_command += ' --show_germline ' if args.print_germline_calls else ""
        indel_fa_predict_command += ' :::: ' + args.output_dir + '/tmp/candidates/INDEL_CANDIDATES_FILES'
        indel_fa_predict_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/7-2_PREDICT_INDEL.log'
        commands_list += [indel_fa_predict_command]

        ## STEP 4: MERGE INDEL VCF
        echo_list.append("[INFO] Merge Full-alignment VCFs")
        indel_fa_mv_command = args.pypy + ' ' + main_entry + ' sort_vcf'
        indel_fa_mv_command += ' --ref_fn ' + args.ref_fn
        indel_fa_mv_command += ' --contigs_fn ' + os.path.join(args.output_dir, 'tmp', 'CONTIGS')
        indel_fa_mv_command += ' --input_dir ' + args.output_dir + '/tmp/vcf_output'
        indel_fa_mv_command += ' --vcf_fn_prefix ' + 'indel_fa_'
        indel_fa_mv_command += ' --output_fn ' + args.output_dir + '/tmp/vcf_output/indel_full_alignment.vcf'
        commands_list += [indel_fa_mv_command]

        indel_pileup_fn = args.output_dir + '/tmp/vcf_output/indel_pileup.vcf'
        indel_fa_fn = args.output_dir + '/tmp/vcf_output/indel_full_alignment.vcf'
        if args.platform != 'ilmn':
            echo_list.append("[INFO] Indel Haplotype filtering")
            indel_hap_g_command = '( ' + time + args.pypy + ' ' + main_entry + ' haplotype_filtering'
            indel_hap_g_command += ' --tumor_bam_fn ' + tumor_bam_prefix
            indel_hap_g_command += ' --ref_fn ' + args.ref_fn
            indel_hap_g_command += ' --germline_vcf_fn ' + clair3_output_path + '/clair3_tumor_output/merge_output.vcf.gz'
            indel_hap_g_command += ' --pileup_vcf_fn ' + args.output_dir + '/tmp/vcf_output/indel_pileup.vcf'
            indel_hap_g_command += ' --full_alignment_vcf_fn ' + args.output_dir + '/tmp/vcf_output/indel_full_alignment.vcf'
            indel_hap_g_command += ' --output_dir ' + args.output_dir + '/tmp/vcf_output'
            indel_hap_g_command += ' --samtools ' + args.samtools
            indel_hap_g_command += ' --pypy3 ' + args.pypy
            indel_hap_g_command += ' --parallel ' + args.parallel
            indel_hap_g_command += ' --threads ' + str(args.threads)
            indel_hap_g_command += ' --debug ' if args.debug else ''
            indel_hap_g_command += ' --show_ref ' if args.print_ref_calls else ''
            indel_hap_g_command += ' --is_indel '
            indel_hap_g_command += ' --apply_post_processing False' if not args.apply_post_processing else ''
            indel_hap_g_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/8_INDEL_HAP_FILTER.log'
            commands_list += [indel_hap_g_command]

            indel_pileup_fn = args.output_dir + '/tmp/vcf_output/indel_pileup_filter.vcf'
            indel_fa_fn = args.output_dir + '/tmp/vcf_output/indel_full_alignment_filter.vcf'

        echo_list.append("[INFO] STEP 8: Merge and sort Indel VCF")
        indel_sort_vcf_command = '( ' + time + args.pypy + ' ' + main_entry + ' merge_vcf'
        indel_sort_vcf_command += ' --ref_fn ' + args.ref_fn
        indel_sort_vcf_command += ' --pileup_vcf_fn ' + indel_pileup_fn
        indel_sort_vcf_command += ' --full_alignment_vcf_fn ' + indel_fa_fn
        indel_sort_vcf_command += ' --output_fn ' + args.output_dir + '/{}.vcf'.format(args.indel_output_prefix)
        indel_sort_vcf_command += ' --platform ' + args.platform
        indel_sort_vcf_command += ' --qual ' + str(args.indel_min_qual)
        indel_sort_vcf_command += ' --sample_name ' + str(args.sample_name)
        indel_sort_vcf_command += ' --enable_indel_calling ' + str(args.enable_indel_calling)
        indel_sort_vcf_command += ' --indel_calling '
        indel_sort_vcf_command += ' --prefer_recall ' + str(args.prefer_recall)
        indel_sort_vcf_command += ' --cmdline ' + args.output_dir + '/tmp/CMD'
        indel_sort_vcf_command += ' ) 2>&1 | tee ' + args.output_dir + '/logs/8_MVI.log'
        commands_list += [indel_sort_vcf_command]

        if args.genotyping_mode_vcf_fn is not None or args.hybrid_mode_vcf_fn is not None:
            echo_list.append("[INFO] Add reference calls to the output VCF output")
            indel_genotyping_command = args.pypy + ' ' + main_entry + ' add_back_missing_variants_in_genotyping'
            indel_genotyping_command += ' --genotyping_mode_vcf_fn ' + str(args.genotyping_mode_vcf_fn)
            indel_genotyping_command += ' --hybrid_mode_vcf_fn ' + str(args.hybrid_mode_vcf_fn)
            indel_genotyping_command += ' --call_fn ' + args.output_dir + '/{}.vcf.gz'.format(args.indel_output_prefix)
            indel_genotyping_command += ' --output_fn ' + args.output_dir + '/{}.vcf'.format(args.indel_output_prefix)
            indel_genotyping_command += ' --candidates_folder ' + args.output_dir + '/tmp/candidates'
            indel_genotyping_command += ' 2>&1 | tee ' + args.output_dir + '/logs/9_GTI.log'
            commands_list += [indel_genotyping_command]

    # excute commands step by step
    skip_steps = args.skip_steps.rstrip().split(',') if args.skip_steps else None
    stdout = sys.stdout if args.tee is None else args.tee.stdin
    for i, (command, echo) in enumerate(zip(commands_list, echo_list)):

        logging(echo)
        logging("[INFO] RUN THE FOLLOWING COMMAND:")
        logging(command)
        logging("")
        if not args.dry_run:
            if skip_steps is not None and str(i+1) in skip_steps:
                logging("[INFO] --skip_steps is enabled, skip running step {}.".format(i+1))
                logging("")
                continue
            try:
                return_code = subprocess.check_call(command, shell=True, stdout=stdout)
            except subprocess.CalledProcessError as e:
                sys.stderr.write("ERROR in STEP {}, THE FOLLOWING COMMAND FAILED: {}\n".format(i+1, command))
                exit(1)
        logging("")

    if args.remove_intermediate_dir:
        logging("[INFO] Removing intermediate files in {}/tmp ...".format(args.output_dir))
        subprocess.run('rm -rf {}/tmp'.format(args.output_dir), shell=True)

    if not args.dry_run and args.enable_clair3_germline_output and args.platform != 'ilmn':
        print("cp {} {}/clair3_tumor_germline_output.vcf.gz".format(clair3_output_path + '/clair3_tumor_output/merge_output.vcf.gz', args.output_dir))
        subprocess.run("cp {} {}/clair3_tumor_germline_output.vcf.gz".format(clair3_output_path + '/clair3_tumor_output/merge_output.vcf.gz', args.output_dir), shell=True,
                       stdout=subprocess.PIPE, stderr=subprocess.PIPE)
        subprocess.run("cp {} {}/clair3_tumor_germline_output.vcf.gz.tbi".format(clair3_output_path + '/clair3_tumor_output/merge_output.vcf.gz.tbi', args.output_dir), shell=True,
                       stdout=subprocess.PIPE, stderr=subprocess.PIPE)
        if not args.normal_vcf_fn:
            subprocess.run("cp {} {}/clair3_normal_germline_output.vcf.gz".format(
                clair3_output_path + '/clair3_normal_output/merge_output.vcf.gz', args.output_dir), shell=True,
                           stdout=subprocess.PIPE, stderr=subprocess.PIPE)
            subprocess.run("cp {} {}/clair3_normal_germline_output.vcf.gz.tbi".format(
                clair3_output_path + '/clair3_normal_output/merge_output.vcf.gz.tbi', args.output_dir), shell=True,
                           stdout=subprocess.PIPE, stderr=subprocess.PIPE)


def somatic_parser():

    parser = argparse.ArgumentParser(
        description="Run ClairS for somatic variant calling. Example run: run_clairs -T TUMOR_BAM -N NORMAL_BAM -R REF -o OUTPUT_DIR -t THREADS -p PLATFORM")

    # print version
    parser.add_argument('-v', '--version', action='version',
                        version='%(prog)s {}'.format(param.version))

    required_params = parser.add_argument_group('Required parameters')
    required_params.add_argument(
        '-T',
        "--tumor_bam_fn",
        type=str,
        required=True,
        default=None,
        help="Tumor BAM file input. The input file must be samtools indexed."
    )

    required_params.add_argument(
        "-N",
        "--normal_bam_fn",
        type=str,
        required=True,
        default=None,
        help="Normal BAM file input. The input file must be samtools indexed."
    )

    required_params.add_argument(
        "-R",
        "--ref_fn",
        type=str,
        required=True,
        default=None,
        help="FASTA reference file input. The input file must be samtools indexed."
    )

    required_params.add_argument(
        "-o",
        "--output_dir",
        type=str,
        required=True,
        default=None,
        help="VCF output directory."
    )

    required_params.add_argument(
        "-t",
        "--threads",
        required=True,
        type=int,
        default=None,
        help="Max #threads to be used."
    )

    required_params.add_argument(
        "-p",
        "--platform",
        required=True,
        type=str,
        default=None,
        help="Select the sequencing platform of the input. Possible options: {ont_r10_dorado_sup_4khz, ont_r10_dorado_sup_5khz_ss, ont_r10_dorado_sup_5khz_ssrs, ont_r10_dorado_hac_5khz, ont_r10_dorado_hac_4khz, ont_r10_guppy, ont_r9_guppy, ilmn, hifi_sequel2, hifi_revio}."
    )

    optional_params = parser.add_argument_group('Optional parameters')
    optional_params.add_argument(
        "-P",
        "--pileup_model_path",
        type=str,
        default=None,
        help="Specify the path to your own somatic calling pileup model."
    )

    optional_params.add_argument(
        "-F",
        "--full_alignment_model_path",
        type=str,
        default=None,
        help="Specify the path to your own somatic calling full-alignment model."
    )

    optional_params.add_argument(
        "-c",
        "--ctg_name",
        type=str,
        default=None,
        help="The name of the contigs to be processed. Split by ',' for multiple contigs. Default: all contigs will be processed."
    )

    region_group = optional_params.add_mutually_exclusive_group(required=False)

    region_group.add_argument(
        "-r",
        "--region",
        type=str,
        default=None,
        help="A region to be processed. Format: `ctg_name:start-end` (start is 1-based)."
    )

    region_group.add_argument(
        "-b",
        "--bed_fn",
        type=str,
        default=None,
        help="Path to a BED file. Call variants only in the provided BED regions."
    )

    region_group.add_argument(
        "-G",
        "--genotyping_mode_vcf_fn",
        type=str,
        default=None,
        help="VCF file input containing candidate sites to be genotyped. Variants will only be called at the sites in the VCF file if provided."
    )

    optional_params.add_argument(
        "-H",
        "--hybrid_mode_vcf_fn",
        type=str,
        default=None,
        help="Enable hybrid calling mode that combines the de novo calling results and genotyping results at the positions in the VCF file given."
    )

    mode_group = optional_params.add_mutually_exclusive_group(required=False)
    mode_group.add_argument(
        '-q',
        "--qual",
        type=int,
        default=None,
        help=SUPPRESS
    )

    optional_params.add_argument(
        "--snv_min_qual",
        type=int,
        default=None,
        help="If set, SNV variants with >SNV_MIN_QUAL will be marked as PASS, or LowQual otherwise."
    )

    optional_params.add_argument(
        "--indel_min_qual",
        type=int,
        default=None,
        help="If set, INDEL variants with >INDEL_MIN_QUAL will be marked as PASS, or LowQual otherwise."
    )

    optional_params.add_argument(
        "--snv_min_af",
        type=float,
        default=None,
        help="Minimal SNV AF required for a variant to be called. Decrease SNV_MIN_AF might increase a bit of sensitivity, but in trade of precision, speed and accuracy. Default: 0.05."
    )

    optional_params.add_argument(
        "--indel_min_af",
        type=float,
        default=None,
        help="Minimal Indel AF required for a variant to be called. Decrease INDEL_MIN_AF might increase a bit of sensitivity, but in trade of precision, speed and accuracy. Default: 0.1 for ONT, 0.05 for other platforms."
    )

    optional_params.add_argument(
        "--indel_max_af_in_normal",
        type=float,
        default=None,
        help="Maximal Indel AF allowed in normal BAM for a called variant If set, variants with NAF<INDEL_MAX_AF_IN_NORMAL will be marked LowQual. Default: None."
    )

    optional_params.add_argument(
        "--min_coverage",
        type=int,
        default=None,
        help="Minimal coverage required for a variant to be called. Default: 4."
    )

    optional_params.add_argument(
        "--chunk_size",
        type=int,
        default=None,
        help="The size of each chuck for parallel processing. Default: 5000000."
    )

    optional_params.add_argument(
        "-s",
        "--sample_name",
        type=str,
        default="SAMPLE",
        help="Define the sample name to be shown in the VCF file. Default: SAMPLE."
    )

    optional_params.add_argument(
        "--output_prefix",
        type=str,
        default="output",
        help="Prefix for output VCF filename. Default: output."
    )

    optional_params.add_argument(
        "--remove_intermediate_dir",
        action='store_true',
        help="Remove intermediate directory before finishing to save disk space."
    )

    optional_params.add_argument(
        "--include_all_ctgs",
        action='store_true',
        help="Call variants on all contigs, otherwise call in chr{1..22} and {1..22}."
    )

    optional_params.add_argument(
        "--print_ref_calls",
        action='store_true',
        help="Show reference calls (0/0) in VCF file."
    )

    optional_params.add_argument(
        "--print_germline_calls",
        action='store_true',
        help="Show germline calls in VCF file."
    )

    optional_params.add_argument(
        '-d',
        "--dry_run",
        action='store_true',
        help="Print the commands that will be ran."
    )

    optional_params.add_argument(
        "--python",
        type=str,
        default="python3",
        help="Absolute path of python, python3 >= 3.9 is required."
    )

    optional_params.add_argument(
        "--pypy",
        type=str,
        default="pypy3",
        help="Absolute path of pypy3, pypy3 >= 3.6 is required."
    )

    optional_params.add_argument(
        "--samtools",
        type=str,
        default="samtools",
        help="Absolute path of samtools, samtools version >= 1.10 is required."
    )

    optional_params.add_argument(
        "--parallel",
        type=str,
        default="parallel",
        help="Absolute path of parallel, parallel >= 20191122 is required."
    )

    # options for advanced users
    optional_params.add_argument(
        "--normal_vcf_fn",
        type=str,
        default=None,
        help="EXPERIMENTAL: Path to normal VCF file with pattern *.vcf.gz. Setting this will skip varaint calling on normal BAM file input"
    )

    halotagging_group = optional_params.add_mutually_exclusive_group(required=False)
    halotagging_group.add_argument(
        "--disable_phasing",
        action='store_true',
        help="EXPERIMENTAL: Disable phasing with longphase or whatshap. Usually leads to significant performance loss."
    )

    halotagging_group.add_argument(
        "--haplotagged_tumor_bam_provided_so_skip_intermediate_phasing_and_haplotagging",
        action='store_true',
        help="EXPERIMENTAL: Use haplotagged tumor bam as input, If enabled, will skip intermediate phasing and haplotagging, and the haplotype information will be acquired from haplotagged tumor BAM. Default: disabled."
    )

    ont_params = parser.add_argument_group('ONT parameters')
    ont_params.add_argument(
        "--enable_indel_calling",
        action='store_true',
        help="EXPERIMENTAL: Enable Indel calling, 'ont_r9_guppy' and 'ilmn' platforms are not supported. The calling time would increase significantly. default: disabled."
    )

    optional_params.add_argument(
        "--enable_clair3_germline_output",
        action='store_true',
        help="EXPERIMENTAL: Use Clair3 default calling settings than Clair3 fast calling setting for tumor and normal germline varaint calling. The calling time would increase ~40 percent, Default: disabled"
    )

    optional_params.add_argument(
        "--enable_verdict",
        action='store_true',
        help="EXPERIMENTAL: Use Verdict to tag the germline variant in CNV regions. We suggest using the parameter only for sample with tumor purity lower than 0.8, Default: disabled"
    )

    ont_params.add_argument(
        "--indel_output_prefix",
        type=str,
        default="indel",
        help="Prefix for Indel output VCF filename. Default: indel."
    )

    ont_params.add_argument(
        "--indel_pileup_model_path",
        type=str,
        default=None,
        help="Specify the path to your own somatic calling indel pileup model."
    )

    ont_params.add_argument(
        "--indel_full_alignment_model_path",
        type=str,
        default=None,
        help="Specify the path to your own somatic calling indel full-alignment model."
    )

    clair3_params = parser.add_argument_group('Clair3 parameters')

    clair3_params.add_argument(
        "--clair3_path",
        type=str_none,
        default=None,
        help='Clair3 main entry path.'
    )

    clair3_params.add_argument(
        "--clair3_model_path",
        type=str,
        default=None,
        help='Clair3 pileup and full-alignment model path.'
    )

    clair3_params.add_argument(
        "--longphase",
        type=str,
        default=None,
        help="Absolute path of longphase, longphase >= 1.7 is required."
    )

    clair3_params.add_argument(
        "--whatshap",
        type=str,
        default=None,
        help="Absolute path of whatshap, whatshap >= 1.0 is required."
    )

    clair3_params.add_argument(
        "--whatshap_for_phasing",
        type=str2bool,
        default=None,
        help="Use whatshap for phasing."
    )

    variant_phasing_group = optional_params.add_mutually_exclusive_group(required=False)

    variant_phasing_group.add_argument(
        "--use_heterozygous_snp_in_normal_sample_and_tumor_bam_for_intermediate_phasing",
        type=str2bool,
        default=None,
        help="EXPERIMENTAL: Use the heterozygous SNPs in normal VCF called by Clair3 and the tumor BAM for intermediate phasing. Option: {True, False}. Default: True."
    )

    variant_phasing_group.add_argument(
        "--use_heterozygous_snp_in_tumor_sample_and_tumor_bam_for_intermediate_phasing",
        type=str2bool,
        default=None,
        help="EXPERIMENTAL: Use the heterozygous SNPs in tumor VCF called by Clair3 and the tumor BAM for intermediate phasing. Option: {True, False}. Default: False."
    )

    variant_phasing_group.add_argument(
        "--use_heterozygous_snp_in_normal_sample_and_normal_bam_for_intermediate_phasing",
        type=str2bool,
        default=None,
        help="EXPERIMENTAL: Use the heterozygous SNPs in normal VCF called by Clair3 and the normal BAM for intermediate phasing. Option: {True, False}. Default: False."
    )

    variant_phasing_group.add_argument(
        "--use_heterozygous_snp_in_tumor_sample_and_normal_bam_for_intermediate_phasing",
        type=str2bool,
        default=None,
        help="EXPERIMENTAL: Use the heterozygous SNPs in tumor VCF called by Clair3 and the normal BAM for intermediate phasing. Option: {True, False}. Default: False."
    )

    optional_params.add_argument(
        "--use_heterozygous_indel_for_intermediate_phasing",
        type=str2bool,
        default=None,
        help="EXPERIMENTAL: Use the heterozygous Indels in normal and tumor VCFs called by Clair3 for intermediate phasing. Option: {True, False}. Default: True."
    )

    optional_params.add_argument(
        "--use_longphase_for_intermediate_haplotagging",
        type=str2bool,
        default=None,
        help="EXPERIMENTAL: Use the longphase instead of whatshap for intermediate haplotagging. Option: {True, False}. Default: True."
    )

    # options for internal process control
    ## Deprecated since v0.3.1. Use the heterozygous SNPs in normal VCF called by Clair3 for intermediate phasing. Option: {True, False}. Default: True.
    variant_phasing_group.add_argument(
        "--use_heterozygous_snp_in_normal_sample_for_intermediate_phasing",
        type=str2bool,
        default=None,
        help=SUPPRESS
    )

    ## Deprecated since v0.3.1. Use the heterozygous SNPs in tumor VCF called by Clair3 for intermediate phasing. Option: {True, False}. Default: False.
    variant_phasing_group.add_argument(
        "--use_heterozygous_snp_in_tumor_sample_for_intermediate_phasing",
        type=str2bool,
        default=None,
        help=SUPPRESS
    )

    ## Use normal BAM instead of tumor BAM for intermediate phasing. Option: {True, False}. Default: False.
    optional_params.add_argument(
        "--use_normal_bam_for_intermediate_phasing",
        type=str2bool,
        default=None,
        help=SUPPRESS
    )

    ## Prefer maximized recall
    mode_group.add_argument(
        "--prefer_recall",
        action='store_true',
        help=SUPPRESS
    )

    ## Prefer balance between precision and recall
    mode_group.add_argument(
        "--prefer_balance",
        action='store_true',
        help=SUPPRESS
    )

    ## Phase normal BAM in calling
    ont_params.add_argument(
        "--phase_normal",
        type=str2bool,
        default=False,
        help=SUPPRESS
    )

    ## Phase tumor BAM in calling
    ont_params.add_argument(
        "--phase_tumor",
        type=str2bool,
        default=None,
        help=SUPPRESS
    )

    #enable parameters for liquid tumor sample
    optional_params.add_argument(
        "--enable_params_for_liquid_tumor_sample",
        action='store_true',
        help=SUPPRESS
    )

    ## Minimum base quality
    optional_params.add_argument(
        "--min_bq",
        type=int,
        default=None,
        help=SUPPRESS
    )

    ##Clair3 options used in calling
    clair3_params.add_argument(
        "--clair3_option",
        type=clair3_option_type,
        default=None,
        help=SUPPRESS
    )

    ## Clair3 germline calling minimum coverage
    optional_params.add_argument(
        "--clair3_min_coverage",
        type=int,
        default=None,
        help=SUPPRESS
    )
    
    ## Clair3 germline snp minimum allele frequency
    optional_params.add_argument(
        "--clair3_snp_min_af",
        type=float,
        default=None,
        help=SUPPRESS
    )
    
    ## Clair3 germline snp minimum allele frequency
    optional_params.add_argument(
        "--clair3_indel_min_af",
        type=float,
        default=None,
        help=SUPPRESS
    )
    
    ##If set then will use GPUs for inference. CUDA required
    optional_params.add_argument(
        "-g",
        "--use_gpu",
        action='store_true',
        help=SUPPRESS
    )

    ##List of (ctg_name, chunk_id, chunk_num)
    optional_params.add_argument(
        "--chunk_list",
        type=str,
        default=None,
        help=SUPPRESS
    )

    ##Apply realignment for short-read data
    optional_params.add_argument(
        "--enable_realignment",
        type=str2bool,
        default=True,
        help=SUPPRESS
    )

    ##Apply haplotype filter for long-read data
    optional_params.add_argument(
        "--apply_post_processing",
        type=str2bool,
        default=True,
        help=SUPPRESS
    )

    optional_params.add_argument(
        "--debug",
        type=str2bool,
        default=False,
        help=SUPPRESS
    )

    optional_params.add_argument(
        "--chunk_num",
        type=int,
        default=None,
        help=SUPPRESS
    )

    optional_params.add_argument(
        "--output_path",
        type=str,
        default=None,
        help=SUPPRESS
    )

    optional_params.add_argument(
        "--cnv_resource_dir",
        type=str,
        default=None,
        help=SUPPRESS
    )

    optional_params.add_argument(
        "--allele_counter_dir",
        type=str,
        default=None,
        help=SUPPRESS
    )

    optional_params.add_argument(
        "--skip_steps",
        type=str,
        default=None,
        help=SUPPRESS
    )

    optional_params.add_argument(
        "--tee",
        type=str,
        default=None,
        help=SUPPRESS
    )

    optional_params.add_argument(
        "--conda_prefix",
        type=str,
        default=None,
        help=SUPPRESS
    )

    optional_params.add_argument(
        "--cmdline",
        type=str,
        default=None,
        help=SUPPRESS
    )

    return parser

def main():
    """
    Main interface for ClairS.
    """

    global args

    call_start_time = time()

    parser = somatic_parser()
    args = parser.parse_args()

    args.output_dir = folder_path_from(args.output_dir, create_not_found=True)
    tee_logger = os.path.join(args.output_dir, 'run_clairs.log' if not args.dry_run else "run_clairs_dry_run.log")
    if os.path.exists(tee_logger):
        subprocess.run("mv {} {}".format(tee_logger, tee_logger + '.bak'), shell=True)
    try:
        args.tee = subprocess.Popen(['tee', tee_logger], stdin=subprocess.PIPE, bufsize=0)
    except:
        logging(log_warning("[WARNING] `tee` not found, disable `tee` logging!"))
        args.tee = None

    logging("")

    args = print_command_line(args)
    args = check_args(args)
    args = print_args(args)
    somatic_calling(args)

    runtime = time() - call_start_time
    logging("[INFO] Total time elapsed: %im%.2fs\n" % (int(runtime/60), int(runtime % 60)))
    if args.enable_verdict and os.path.exists("{}/{}_cnv_germline_tagged.vcf.gz\n".format(args.output_dir, args.output_prefix)):
        logging("[INFO] Finish calling, output file: {}/{}_cnv_germline_tagged.vcf.gz\n".format(args.output_dir, args.output_prefix))
    else:
        logging("[INFO] Finish calling, output file: {}/{}.vcf.gz\n".format(args.output_dir, args.output_prefix))
    if args.enable_indel_calling:
        logging("[INFO] Finish calling, snv output file: {}/snv.vcf.gz\n".format(args.output_dir))
        logging("[INFO] Finish calling, indel output file: {}/{}.vcf.gz\n".format(args.output_dir, args.indel_output_prefix))

    if args.enable_clair3_germline_output and args.platform != 'ilmn':
        logging("[INFO] Finish calling, Clair3 tumor BAM germline output file: {}/clair3_tumor_germline_output.vcf.gz\n".format(args.output_dir))
        if not args.normal_vcf_fn:
            logging("[INFO] Finish calling, Clair3 normal BAM germline output file: {}/clair3_normal_germline_output.vcf.gz\n".format(args.output_dir))

    if args.tee is not None:
        args.tee.stdin.close()

if __name__ == '__main__':
    main()