def __call__(self, args):
print(f'Database building started at {timestamp()}.')
# load configurations
self.cfg = load_configs()
# read and validate arguments
self.set_parameters(args)
if not self.skipdownload:
# connect to NCBI FTP server
self.connect_server()
# create target directory
makedirs(self.output, exist_ok=True)
# retrieve taxonomy database
self.retrieve_taxdump()
# retrieve genome assembly summary
self.retrieve_summary()
# retrieve genome categories
self.retrieve_categories()
# filter genomes
self.filter_genomes()
# identify taxonomy of genomes
self.identify_taxonomy()
# sample genomes by taxonomy
self.sample_by_taxonomy()
# download genomes
self.download_genomes()
# extract genomes
self.extract_genomes()
# identify genome lineages
self.genome_lineages()
# write genome metadata
self.genome_metadata()
# build taxonomy database
self.build_taxdump()
# build protein-to-taxonomy map
self.build_taxonmap()
# build protein sequence database
self.compile_database()
# clean up
if hasattr(self, 'mkdtemp'):
rmtree(self.tmpdir)
print(f'Database building finished at {timestamp()}.')
def test_set_parameters(self):
me = Analyze()
me.cfg = load_configs()
def args():
return None
# input is file
infile = join(self.datadir, 'DnaK', 'search', 'sample.tsv')
outdir = join(self.tmpdir, 'output')
args.input = infile
args.output = outdir
args.noise = 0.75
me.set_parameters(args)
self.assertEqual(me.input, infile)
self.assertEqual(me.output, outdir)
self.assertTrue(isdir(outdir))
self.assertDictEqual(me.input_map, {'sample': infile})
self.assertEqual(me.noise, 75)
# coverage threshold too small
args.input_cov = 25
with self.assertRaises(ValueError) as ctx:
me.set_parameters(args)
msg = 'Taxonomy coverage for auto-interence must be at least 50%.'
self.assertEqual(str(ctx.exception), msg)
args.input_cov = 75
# input is directory
indir = join(self.datadir, 'DnaK', 'search')
args.input = indir
me.set_parameters(args)
self.assertEqual(me.input, indir)
self.assertDictEqual(me.input_map, {'sample': infile})
rmtree(outdir)
# input is invalid
not_path = 'I am not a path'
args.input = not_path
with self.assertRaises(ValueError) as ctx:
me.set_parameters(args)
msg = f'Invalid input data file or directory: {not_path}.'
self.assertEqual(str(ctx.exception), msg)
# input has no search result
args.input = self.tmpdir
with self.assertRaises(ValueError) as ctx:
me.set_parameters(args)
msg = f'No input data are found under: {self.tmpdir}.'
self.assertEqual(str(ctx.exception), msg)
# no input (which is okay)
delattr(me, 'input_map')
args.input = None
me.set_parameters(args)
self.assertFalse(hasattr(me, 'input_map'))
def __call__(self, args):
print('Analysis started at {}.'.format(timestamp()))
# load configurations
self.cfg = load_configs()
# read and validate arguments
self.set_parameters(args)
# use existing score table
score_file = join(self.output, 'scores.tsv')
if self.from_scores and isfile(score_file):
self.df = pd.read_csv(score_file, sep='\t')
else:
# read input taxonomy and search results
self.read_input()
# assign taxonomy to input genomes
self.assign_taxonomy()
# define the three groups for search hits
self.define_groups()
# calculate scores for the three groups per protein
self.calc_scores()
# generate a table of calculated scores
self.make_score_table()
# no distal hits
if not self.df['distal'].any():
print('WARNING: No hit is assigned to distal group. Cannot '
'predict HGTs.')
return
# remove orphans
if not self.orphans:
self.remove_orphans()
# remove outliers
if self.outliers != 'none':
self.remove_outliers()
# predict HGTs
self.predict_hgt()
print('Analysis finished at {}.'.format(timestamp()))
def test_load_configs(self):
obs = load_configs()
self.assertIn('database', obs)