Mercurial > repos > kkonganti > cfsan_bettercallsal

annotate 0.6.1/bin/dl_pdg_metadata.py @ 14:b0a37e88ecb5

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
"planemo upload"
author kkonganti
date Thu, 07 Sep 2023 10:13:31 -0400
parents 749faef1caa9
children
rev   line source
kkonganti@11 1 #!/usr/bin/env python3
kkonganti@11 2
kkonganti@11 3 # Kranti Konganti
kkonganti@11 4
kkonganti@11 5 import os
kkonganti@11 6 import shutil
kkonganti@11 7 import tempfile
kkonganti@11 8 import argparse
kkonganti@11 9 import inspect
kkonganti@11 10 import logging
kkonganti@11 11 import re
kkonganti@11 12 from urllib.request import urlopen
kkonganti@11 13 from html.parser import HTMLParser
kkonganti@11 14
kkonganti@11 15 # Set logging.f
kkonganti@11 16 logging.basicConfig(
kkonganti@11 17 format="\n" + "=" * 55 + "\n%(asctime)s - %(levelname)s\n" + "=" * 55 + "\n%(message)s\n",
kkonganti@11 18 level=logging.DEBUG,
kkonganti@11 19 )
kkonganti@11 20
kkonganti@11 21 # Multiple inheritence for pretty printing of help text.
kkonganti@11 22 class MultiArgFormatClasses(argparse.RawTextHelpFormatter, argparse.ArgumentDefaultsHelpFormatter):
kkonganti@11 23 pass
kkonganti@11 24
kkonganti@11 25
kkonganti@11 26 # HTMLParser override class to get PDG release and latest Cluster .tsv file
kkonganti@11 27 class NCBIPathogensHTMLParser(HTMLParser):
kkonganti@11 28 def __init__(self, *, convert_charrefs: bool = ...) -> None:
kkonganti@11 29 super().__init__(convert_charrefs=convert_charrefs)
kkonganti@11 30 self.reset()
kkonganti@11 31 self.href_data = list()
kkonganti@11 32
kkonganti@11 33 def handle_data(self, data):
kkonganti@11 34 self.href_data.append(data)
kkonganti@11 35
kkonganti@11 36
kkonganti@11 37 def dl_pdg(**kwargs) -> None:
kkonganti@11 38 """
kkonganti@11 39 Method to save the PDG metadata file and
kkonganti@11 40 return the latest PDG release.
kkonganti@11 41 """
kkonganti@11 42 db_path, url, regex, suffix, overwrite, release = [kwargs[k] for k in kwargs.keys()]
kkonganti@11 43
kkonganti@11 44 if (db_path or url) == None:
kkonganti@11 45 logging.error("Please provide absolute UNIX path\n" + "to store the result DB flat files.")
kkonganti@11 46 exit(1)
kkonganti@11 47
kkonganti@11 48 if re.match(r"^PDG\d+\.\d+$", release):
kkonganti@11 49 url = re.sub("latest_snps", release.strip(), url)
kkonganti@11 50
kkonganti@11 51 html_parser = NCBIPathogensHTMLParser()
kkonganti@11 52 logging.info(f"Finding latest NCBI PDG release at:\n{url}")
kkonganti@11 53
kkonganti@11 54 with urlopen(url) as response:
kkonganti@11 55 with tempfile.NamedTemporaryFile(delete=False) as tmp_html_file:
kkonganti@11 56 shutil.copyfileobj(response, tmp_html_file)
kkonganti@11 57
kkonganti@11 58 with open(tmp_html_file.name, "r") as html:
kkonganti@11 59 html_parser.feed("".join(html.readlines()))
kkonganti@11 60
kkonganti@11 61 pdg_filename = re.search(regex, "".join(html_parser.href_data)).group(0)
kkonganti@11 62 pdg_release = pdg_filename.rstrip(suffix)
kkonganti@11 63 pdg_metadata_url = "/".join([url, pdg_filename])
kkonganti@11 64 pdg_release = pdg_filename.rstrip(suffix)
kkonganti@11 65 dest_dir = os.path.join(db_path, pdg_release)
kkonganti@11 66
kkonganti@11 67 logging.info(f"Found NCBI PDG file:\n{pdg_metadata_url}")
kkonganti@11 68
kkonganti@11 69 if (
kkonganti@11 70 not overwrite
kkonganti@11 71 and re.match(r".+?\.metadata\.tsv$", pdg_filename)
kkonganti@11 72 and os.path.exists(dest_dir)
kkonganti@11 73 ):
kkonganti@11 74 logging.error(f"DB path\n{dest_dir}\nalready exists. Please use -f to overwrite.")
kkonganti@11 75 exit(1)
kkonganti@11 76 elif overwrite and not re.match(r".+?\.reference_target\.cluster_list\.tsv$", pdg_filename):
kkonganti@11 77 shutil.rmtree(dest_dir, ignore_errors=True) if os.path.exists(dest_dir) else None
kkonganti@11 78 os.makedirs(dest_dir)
kkonganti@11 79 elif (
kkonganti@11 80 not overwrite
kkonganti@11 81 and re.match(r".+?\.metadata\.tsv$", pdg_filename)
kkonganti@11 82 and not os.path.exists(dest_dir)
kkonganti@11 83 ):
kkonganti@11 84 os.makedirs(dest_dir)
kkonganti@11 85
kkonganti@11 86 tsv_at = os.path.join(dest_dir, pdg_filename)
kkonganti@11 87 logging.info(f"Saving to:\n{tsv_at}")
kkonganti@11 88
kkonganti@11 89 with urlopen(pdg_metadata_url) as response:
kkonganti@11 90 with open(tsv_at, "w") as tsv:
kkonganti@11 91 tsv.writelines(response.read().decode("utf-8"))
kkonganti@11 92
kkonganti@11 93 html.close()
kkonganti@11 94 tmp_html_file.close()
kkonganti@11 95 os.unlink(tmp_html_file.name)
kkonganti@11 96 tsv.close()
kkonganti@11 97 response.close()
kkonganti@11 98
kkonganti@11 99 return tsv_at, dest_dir
kkonganti@11 100
kkonganti@11 101
kkonganti@11 102 def main() -> None:
kkonganti@11 103 """
kkonganti@11 104 This script is part of the `bettercallsal_db` Nextflow workflow and is only
kkonganti@11 105 tested on POSIX sytems.
kkonganti@11 106 It:
kkonganti@11 107 1. Downloads the latest NCBI Pathogens Release metadata file, which
kkonganti@11 108 looks like PDGXXXXXXXXXX.2504.metadata.csv and also the SNP cluster
kkonganti@11 109 information file which looks like PDGXXXXXXXXXX.2504.reference_target.cluster_list.tsv
kkonganti@11 110 2. Generates a new metadata file with only required information such as
kkonganti@11 111 computed_serotype, isolates GenBank or RefSeq downloadable genome FASTA
kkonganti@11 112 URL.
kkonganti@11 113 """
kkonganti@11 114
kkonganti@11 115 prog_name = os.path.basename(inspect.stack()[0].filename)
kkonganti@11 116
kkonganti@11 117 parser = argparse.ArgumentParser(
kkonganti@11 118 prog=prog_name, description=main.__doc__, formatter_class=MultiArgFormatClasses
kkonganti@11 119 )
kkonganti@11 120
kkonganti@11 121 # required = parser.add_argument_group("required arguments")
kkonganti@11 122
kkonganti@11 123 parser.add_argument(
kkonganti@11 124 "-db",
kkonganti@11 125 dest="db_path",
kkonganti@11 126 default=os.getcwd(),
kkonganti@11 127 required=False,
kkonganti@11 128 help="Absolute UNIX path to a path where all results files are\nstored.",
kkonganti@11 129 )
kkonganti@11 130 parser.add_argument(
kkonganti@11 131 "-f",
kkonganti@11 132 dest="overwrite_db",
kkonganti@11 133 default=False,
kkonganti@11 134 required=False,
kkonganti@11 135 action="store_true",
kkonganti@11 136 help="Force overwrite a PDG release directory at DB path\nmentioned with -db.",
kkonganti@11 137 )
kkonganti@11 138 parser.add_argument(
kkonganti@11 139 "-org",
kkonganti@11 140 dest="organism",
kkonganti@11 141 default="Salmonella",
kkonganti@11 142 required=False,
kkonganti@11 143 help="The organism to create the DB flat files\nfor.",
kkonganti@11 144 )
kkonganti@11 145 parser.add_argument(
kkonganti@11 146 "-rel",
kkonganti@11 147 dest="release",
kkonganti@11 148 default=False,
kkonganti@11 149 required=False,
kkonganti@11 150 help="If you get a 404 error, try mentioning the actual release identifier.\n"
kkonganti@11 151 + "Ex: For Salmonella, you can get the release identifier by going to:\n"
kkonganti@11 152 + " https://ftp.ncbi.nlm.nih.gov/pathogen/Results/Salmonella\n"
kkonganti@11 153 + "Ex: If you want metadata beloginging to release PDG000000002.2507, then you\n"
kkonganti@11 154 + " would use this command-line option as:\n -rel PDG000000002.2507",
kkonganti@11 155 )
kkonganti@11 156
kkonganti@11 157 args = parser.parse_args()
kkonganti@11 158 db_path = args.db_path
kkonganti@11 159 org = args.organism
kkonganti@11 160 overwrite = args.overwrite_db
kkonganti@11 161 release = args.release
kkonganti@11 162 ncbi_pathogens_loc = "/".join(
kkonganti@11 163 ["https://ftp.ncbi.nlm.nih.gov/pathogen/Results", org, "latest_snps"]
kkonganti@11 164 )
kkonganti@11 165
kkonganti@11 166 if not db_path:
kkonganti@11 167 db_path = os.getcwd()
kkonganti@11 168
kkonganti@11 169 # Save metadata
kkonganti@11 170 file, dest_dir = dl_pdg(
kkonganti@11 171 db_path=db_path,
kkonganti@11 172 url="/".join([ncbi_pathogens_loc, "Metadata"]),
kkonganti@11 173 regex=re.compile(r"PDG\d+\.\d+\.metadata\.tsv"),
kkonganti@11 174 suffix=".metadata.tsv",
kkonganti@11 175 overwrite=overwrite,
kkonganti@11 176 release=release,
kkonganti@11 177 )
kkonganti@11 178
kkonganti@11 179 # Save cluster to target mapping
kkonganti@11 180 dl_pdg(
kkonganti@11 181 db_path=db_path,
kkonganti@11 182 url="/".join([ncbi_pathogens_loc, "Clusters"]),
kkonganti@11 183 regex=re.compile(r"PDG\d+\.\d+\.reference_target\.cluster_list\.tsv"),
kkonganti@11 184 suffix="reference_target\.cluster_list\.tsv",
kkonganti@11 185 overwrite=overwrite,
kkonganti@11 186 release=release,
kkonganti@11 187 )
kkonganti@11 188
kkonganti@11 189 # Create accs.txt for dataformat to fetch required ACC fields
kkonganti@11 190 accs_file = os.path.join(dest_dir, "accs_all.txt")
kkonganti@11 191 with open(file, "r") as pdg_metadata_fh:
kkonganti@11 192 with open(accs_file, "w") as accs_fh:
kkonganti@11 193 for line in pdg_metadata_fh.readlines():
kkonganti@11 194 if re.match(r"^#", line) or line in ["\n", "\n\r", "\r"]:
kkonganti@11 195 continue
kkonganti@11 196 cols = line.strip().split("\t")
kkonganti@11 197 asm_acc = cols[9]
kkonganti@11 198 accs_fh.write(f"{asm_acc}\n") if (asm_acc != "NULL") else None
kkonganti@11 199 accs_fh.close()
kkonganti@11 200 pdg_metadata_fh.close()
kkonganti@11 201
kkonganti@11 202 logging.info("Finished writing accessions for dataformat tool.")
kkonganti@11 203
kkonganti@11 204
kkonganti@11 205 if __name__ == "__main__":
kkonganti@11 206 main()