def __init__(self, top_level_directory, reference_fname, min_length, min_hitidentity, min_hitlength, max_evalue, batch_size, min_alignidentity, min_alignoverlap) :
self.directory = join(top_level_directory, 'alignments')
self.min_length = min_length # glutton
self.min_hitidentity = min_hitidentity # blast
self.min_hitlength = min_hitlength # blast
self.max_evalue = max_evalue # blast
self.min_alignidentity = min_alignidentity # pagan
self.min_alignoverlap = min_alignoverlap # pagan
check_dir(self.directory, create=True)
self.search = All_vs_all_search(batch_size)
self.cleanup_files = []
self.q = None
self.lock = threading.Lock()
self.complete_jobs = 0
self.total_jobs = 0
self.log = get_log()
self.param = GluttonParameters(top_level_directory)
self.db = GluttonDB(reference_fname)
self.param.set_reference(self.db)
self.resume = self.param.able_to_resume()
self.info = GluttonInformation(self.directory, self.param, self.db, resume=self.resume)
self.param.set_full_checksum()
def copy(self, src) :
data_dir = join(self.directory, 'data')
check_dir(data_dir, create=True)
dst = join(data_dir, basename(src))
self.log.info("cp %s %s" % (src, dst))
shutil.copy(src, dst)
return join('data', basename(src))
def __init__(self, project_dir, create=False) :
self.directory = project_dir
self.create = create
check_dir(self.directory, create=self.create)
self.log = get_log()
self.params = self.load(self.parameter_filename)
if not self.params :
self.params = { 'db_species' : None,
'db_release' : None,
'db_filename' : None,
'db_checksum' : None,
'full_checksum' : None,
'sample_checksum' : None,
'samples' : {} }
def __init__(self, alignments_dir, parameters, db, resume=True) :
self.directory = alignments_dir
self.params = parameters
self.db = db
check_dir(self.directory)
self.log = get_log()
self.lock = threading.RLock() # a single function requires this be an RLock over a Lock
# the alignment procedure can take a long time, so everything needs to be
# restartable, in addition - if we restart it then we need to be sure that
# the parameters used are the same, i.e.: same reference database etc etc
self.contig_query_map = {} # file id -> contig id -> query id (file id is provided by the user, called a 'label')
self.query_gene_map = {} # query id -> (gene id, +/-) or None
self.genefamily_filename_map = {} # gene family id -> filename
if resume :
self.read_progress_files()
def __init__(self, top_level_directory, reference_fname, assembler_name, protein_identity, alignment_length, min_gene_coverage, do_not_trim=False, testmode='none') :
self.alignments_dir = join(top_level_directory, 'alignments')
self.output_dir = join(top_level_directory, 'postprocessing')
self.protein_identity = protein_identity
self.alignment_length = alignment_length
self.min_gene_coverage = min_gene_coverage
self.trim = not do_not_trim
self.testmode = testmode
self.scaffold_dir = join(self.output_dir, 'scaffolds')
self.genefamily_msa_dir = join(self.output_dir, 'genefamily_msa')
self.gene_msa_dir = join(self.output_dir, 'gene_msa')
check_dir(self.output_dir, create=True)
check_dir(self.scaffold_dir, create=True)
check_dir(self.genefamily_msa_dir, create=True)
check_dir(self.gene_msa_dir, create=True)
self.log = get_log()
self.param = GluttonParameters(top_level_directory)
self.db = GluttonDB(reference_fname)
self.info = GluttonInformation(self.alignments_dir, self.param, self.db)
# check reference was the same
if not self.param.same_reference(self.db) :
self.log.error("current reference is %s, alignments were performed using %s" % (reference_fname, self.db.filename))
exit(1);
# perhaps slightly overambitious to exit, just stick to a warning
pending,failures = self.info.num_alignments_not_done()
if pending != 0 :
self.log.warn("%d alignments were not run!" % pending)
self.assembler = AssemblerOutput(assembler_name)
# e.g. query39806_orf1
self.orfname_regex = re.compile("^(query\d+)\_orf(\d)$")
def generic_options(args) :
if hasattr(args, 'database_host') and args.database_host :
custom_database(args.database_host,
args.database_port,
args.database_user,
args.database_password)
# threads
if hasattr(args, 'threads') :
set_threads(args.threads)
# tmpdir
if hasattr(args, 'tmpdir') :
try :
check_dir(args.tmpdir, create=True)
set_tmpdir(args.tmpdir)
except OSError :
print >> stderr, "ERROR: %s does not exist..." % args.tmpdir
exit(1)
# gltfile
if hasattr(args, 'gltfile') and args.gltfile :
if not os.path.isfile(args.gltfile) :
print >> stderr, "ERROR: %s does not exist..." % args.gltfile
exit(1)
args.output = args.gltfile
if hasattr(args, 'gltfile') and hasattr(args, 'species') :
if not args.gltfile and not args.species :
print >> stderr, "ERROR: you must specify either the species or an existing GLT file..."
exit(1)
if hasattr(args, 'reference') and args.reference :
if not os.path.isfile(args.reference) :
print >> stderr, "ERROR: %s does not exist..." % args.reference
exit(1)
# verbosity
if hasattr(args, 'verbose') :
set_verbosity(args.verbose)
# ensembl download method
if hasattr(args, 'method') :
set_ensembl_download_method(args.method)
# build does not work for ensembl-genomes if the method is "sql"
if hasattr(args, 'method') and args.method == "sql" :
if args.database != 'ensembl' :
print >> stderr, "ERROR: 'sql' does not work for Ensembl-genomes, use 'biomart' instead..."
exit(1)
# contigs, labels, species, bam files
# contigs, labels amd bam files must be unique
# apart from 'FAKE' bam file
# length of contigs, label, species must be the same
# length of bam files is either zero or same as contigs
if hasattr(args, 'contigs') :
if args.contigs :
if not args.sample or not args.species :
print >> stderr, "ERROR: you must specify one --sample and one --species argument per --contigs file!"
exit(1)
if hasattr(args, 'setupcmd') :
if args.setupcmd == 'add' and None in (args.sample, args.contigs, args.species) :
print >> stderr, "ERROR: to add a sample, you must specify sample identifier, contig FASTA file and species name (see glutton setup -h)"
exit(1)
elif args.setupcmd == 'remove' and not args.sample :
print >> stderr, "ERROR: no sample identifier found"
exit(1)
setup_logging()