def taxon_profile(self, options):
"""Call genes command"""
make_sure_path_exists(options.output_dir)
check_file_exists(options.scaffold_stats_file)
check_file_exists(options.taxonomy_file)
check_file_exists(options.db_file)
gene_files = self._genome_files(options.genome_prot_dir,
options.protein_ext)
if not self._check_protein_seqs(gene_files):
self.logger.warning('All files must contain amino acid sequences.')
sys.exit()
# build gene profile
taxon_profile = TaxonProfile(options.cpus, options.output_dir)
taxon_profile.run(gene_files, options.scaffold_stats_file,
options.db_file, options.taxonomy_file,
options.per_to_classify, options.evalue,
options.per_identity, options.per_aln_len,
options.tmpdir)
self.logger.info('Results written to: %s' % options.output_dir)
def taxon_filter(self, options):
"""Taxon filter command"""
make_sure_path_exists(options.taxon_profile_dir)
# build gene profile
taxon_profile = TaxonProfile(options.cpus, options.taxon_profile_dir)
if False:
taxon_profile.filter(options.genome_threshold,
options.min_scaffold_agreement,
options.max_scaffold_disagreement,
options.min_classified_per,
options.output_file)
else:
taxon_profile.filter(
options.consensus_taxon, options.trusted_scaffold,
options.common_taxa, options.congruent_scaffold,
options.min_classified_per, options.min_classified,
options.consensus_scaffold, options.output_file)
self.logger.info('Results written to: %s' % options.output_file)
def erroneous(self, ssu_hits, ssu_classifications, taxon_profile_dir,
common_taxon_threshold, ssu_min_length, ssu_domain,
ssu_phylum, ssu_class, ssu_order, ssu_family, ssu_genus,
output_dir):
"""Identify scaffolds with SSU genes that have divergent taxonomic classifcation.
Parameters
----------
ssu_hits : d[genome_id][scaffold_id] -> hit information
Hits to SSU genes in each genome.
ssu_classifications : d[genome_id][scaffold_id] -> hit information
File with taxonomic classifications for SSU genes.
taxon_profile_dir : str
common_taxon_threshold : float
ssu_min_length : int
ssu_domain : float
ssu_phylum : float
ssu_class : float
ssu_order : float
ssu_family : float
ssu_genus : float
output_dir : str
Directory for output files.
"""
header = 'Scaffold id\tGenome id\tGenome classification\tIncongruent common taxa set'
header += '\tNo. 16S in Genome'
header += '\t16S Classification\t16S length\t16S e-value\t16S alignment length\t16S percent identity'
header += '\tScaffold length (bp)\n'
fout = open(os.path.join(output_dir, 'ssu_erroneous.tsv'), 'w')
fout.write(header)
taxon_profile = TaxonProfile(1, taxon_profile_dir)
common_taxa = taxon_profile.common_taxa(common_taxon_threshold, 25.0)
genome_taxonomy = taxon_profile.read_genome_taxonomy()
for genome_id, scaffold_ids in ssu_hits.iteritems():
# **** HACK for SRA processing
gid = genome_id.replace('.filtered', '')
for scaffold_id in scaffold_ids:
hmm_model, evalue, _start, _stop, ssu_length, _rev_comp, scaffold_len = ssu_hits[
genome_id][scaffold_id]
evalue = float(evalue)
ssu_length = int(ssu_length)
scaffold_len = int(scaffold_len)
if ssu_length < ssu_min_length:
continue
if scaffold_id not in ssu_classifications[genome_id]:
continue
ssu_taxonomy, _, _, _, per_ident = ssu_classifications[
genome_id][scaffold_id]
per_ident = float(per_ident)
# check if taxa indicated by 16S rRNA gene are congruent with the list of common taxa in genome
ssu_taxonomy = ssu_taxonomy.split(';')
congruent = True
incongruent_common_taxa = None
for r, value in enumerate([
ssu_domain, ssu_phylum, ssu_class, ssu_order,
ssu_family, ssu_genus
]):
if r not in common_taxa[gid]:
# insufficient classified genes to determine common taxa at rank
break
if len(common_taxa[gid][r]) == 0:
# no consistent taxonomic signal at rank so do not filter
break
if ssu_taxonomy[r] == Taxonomy.rank_prefixes[r]:
break
if per_ident < value:
break
common_taxa_at_rank = common_taxa[gid][r]
if ssu_taxonomy[r] not in common_taxa_at_rank:
congruent = False
incongruent_common_taxa = common_taxa_at_rank
break
# report outliers
if not congruent:
row = '%s\t%s\t%s' % (scaffold_id, genome_id,
genome_taxonomy[gid])
row += '\t%s\t%d' % (';'.join(
sorted(
list(incongruent_common_taxa))), len(scaffold_ids))
row += '\t%s\t%s\t%s\t%s\t%s' % tuple(
ssu_classifications[genome_id][scaffold_id])
row += '\t%d\n' % scaffold_len
fout.write(row)
fout.close()