def test_run_best_match_not_in_cluster(self):
'''Test full run where there is a match in cluster, but better match to seq not in cluster'''
all_ref_fasta = os.path.join(data_dir, 'ref_seq_chooser_full_run_best_match_not_in_cluster.allrefs.fa')
cluster_fasta = os.path.join(data_dir, 'ref_seq_chooser_full_run_best_match_not_in_cluster.clusterrefs.fa')
contig_fasta = os.path.join(data_dir, 'ref_seq_chooser_full_run_best_match_not_in_cluster.contigs.fa')
tmp_out = 'tmp.ref_seq_chooser_full_run_best_match_not_in_cluster.fa'
refchooser = ref_seq_chooser.RefSeqChooser(cluster_fasta, all_ref_fasta, contig_fasta, tmp_out, sys.stdout)
refchooser.run()
self.assertEqual('ref2', refchooser.closest_ref_from_all_refs)
self.assertFalse(refchooser.closest_ref_is_in_cluster)
self.assertFalse(os.path.exists(tmp_out))
def test_run_no_nucmer_match(self):
'''Test full run when there is nearest match in cluster, but no nucmer matches'''
all_ref_fasta = os.path.join(data_dir, 'ref_seq_chooser_full_run_no_nucmer_match.allrefs.fa')
cluster_fasta = os.path.join(data_dir, 'ref_seq_chooser_full_run_no_nucmer_match.clusterrefs.fa')
contig_fasta = os.path.join(data_dir, 'ref_seq_chooser_full_run_no_nucmer_match.contigs.fa')
tmp_out = 'tmp.ref_seq_chooser_full_run_no_nucmer_match.fa'
refchooser = ref_seq_chooser.RefSeqChooser(cluster_fasta, all_ref_fasta, contig_fasta, tmp_out, sys.stdout)
refchooser.run()
self.assertEqual(None, refchooser.closest_ref_from_all_refs)
self.assertFalse(refchooser.closest_ref_is_in_cluster)
self.assertFalse(os.path.exists(tmp_out))
def test_run_contained_ref_seq(self):
'''Test full run where ref seq completely contains another seq outside cluster'''
all_ref_fasta = os.path.join(data_dir, 'ref_seq_chooser_full_run_contained_ref_seq.all_refs.fa')
cluster_fasta = os.path.join(data_dir, 'ref_seq_chooser_full_run_contained_ref_seq.cluster_refs.fa')
contig_fasta = os.path.join(data_dir, 'ref_seq_chooser_full_run_contained_ref_seq.contigs.fa')
tmp_out = 'tmp.ref_seq_chooser_full_run_contained_ref_seq.fa'
refchooser = ref_seq_chooser.RefSeqChooser(cluster_fasta, all_ref_fasta, contig_fasta, tmp_out, sys.stdout)
refchooser.run()
self.assertEqual('ref2', refchooser.closest_ref_from_all_refs)
self.assertTrue(refchooser.closest_ref_is_in_cluster)
self.assertTrue(os.path.exists(tmp_out))
os.unlink(tmp_out)
def test_run_best_match_is_in_cluster(self):
'''Test full run where the best match is in the cluster'''
all_ref_fasta = os.path.join(data_dir, 'ref_seq_chooser_full_run_best_match_is_in_cluster.allrefs.fa')
cluster_fasta = os.path.join(data_dir, 'ref_seq_chooser_full_run_best_match_is_in_cluster.clusterrefs.fa')
contig_fasta = os.path.join(data_dir, 'ref_seq_chooser_full_run_best_match_is_in_cluster.contigs.fa')
tmp_out = 'tmp.ref_seq_chooser_full_run_best_match_is_in_cluster.fa'
refchooser = ref_seq_chooser.RefSeqChooser(cluster_fasta, all_ref_fasta, contig_fasta, tmp_out, sys.stdout)
refchooser.run()
self.assertEqual('ref1', refchooser.closest_ref_from_all_refs)
self.assertTrue(refchooser.closest_ref_is_in_cluster)
self.assertTrue(os.path.exists(tmp_out))
os.unlink(tmp_out)
def test_run_flanking_different(self):
'''Test full run where amount of flanking seq varies'''
all_ref_fasta = os.path.join(data_dir, 'ref_seq_chooser_test_flanking.all_refs.fa')
cluster_fasta = os.path.join(data_dir, 'ref_seq_chooser_test_flanking.cluster_refs.fa')
contig_fasta = os.path.join(data_dir, 'ref_seq_chooser_test_flanking.contigs.fa')
expected_fa = os.path.join(data_dir, 'ref_seq_chooser_test_flanking.expected_contigs.fa')
tmp_out = 'tmp.ref_seq_chooser_test_flanking.fa'
refchooser = ref_seq_chooser.RefSeqChooser(cluster_fasta, all_ref_fasta, contig_fasta, tmp_out, sys.stdout)
refchooser.run()
self.assertEqual('ref1', refchooser.closest_ref_from_all_refs)
self.assertTrue(refchooser.closest_ref_is_in_cluster)
self.assertTrue(filecmp.cmp(expected_fa, tmp_out, shallow=False))
os.unlink(tmp_out)
def run(self):
if self.assembler == 'fermilite':
self._assemble_with_fermilite()
elif self.assembler == "spades":
self._assemble_with_spades()
print('Finished running assemblies', flush=True, file=self.log_fh)
self.sequences = {}
# double-check we got some contigs
number_of_contigs = pyfastaq.tasks.count_sequences(
self.all_assembly_contigs_fa) if os.path.exists(
self.all_assembly_contigs_fa) else 0
if number_of_contigs == 0:
self.assembled_ok = False
# This is to make this object picklable, to keep multithreading happy
self.log_fh = None
return
else:
self.assembled_ok = True
if self.assembled_ok:
ref_chooser = ref_seq_chooser.RefSeqChooser(
self.ref_fastas,
self.all_reference_fasta,
self.all_assembly_contigs_fa,
self.best_assembly_fa,
self.log_fh,
nucmer_min_id=self.nucmer_min_id,
nucmer_min_len=self.nucmer_min_len,
nucmer_breaklen=self.nucmer_breaklen,
)
ref_chooser.run()
if ref_chooser.closest_ref_from_all_refs is None:
print('Could not find match to reference sequences',
file=self.log_fh)
self.ref_seq_name = None
self.log_fh = None
return
elif not ref_chooser.closest_ref_is_in_cluster:
print('Closest reference',
ref_chooser.closest_ref_from_all_refs,
'was not in cluster',
file=self.log_fh)
self.ref_seq_name = None
self.log_fh = None
return
else:
assert ref_chooser.closest_ref_from_all_refs is not None
self.ref_seq_name = ref_chooser.closest_ref_from_all_refs
print('Closest reference sequence:',
self.ref_seq_name,
file=self.log_fh)
file_reader = pyfastaq.sequences.file_reader(self.ref_fastas)
for ref_seq in file_reader:
if self.ref_seq_name == ref_seq.id:
f_out = pyfastaq.utils.open_file_write(self.ref_fasta)
print(ref_seq, file=f_out)
pyfastaq.utils.close(f_out)
break
contigs_both_strands = self._fix_contig_orientation(
self.best_assembly_fa,
self.ref_fasta,
self.final_assembly_fa,
min_id=self.nucmer_min_id,
min_length=self.nucmer_min_len,
breaklen=self.nucmer_breaklen)
self.has_contigs_on_both_strands = len(contigs_both_strands) > 0
pyfastaq.tasks.file_to_dict(self.final_assembly_fa, self.sequences)
mapping.run_bowtie2(
self.reads1,
self.reads2,
self.final_assembly_fa,
self.final_assembly_bam[:-4],
threads=self.threads,
sort=True,
bowtie2=self.extern_progs.exe('bowtie2'),
bowtie2_version=self.extern_progs.version('bowtie2'),
verbose=True,
verbose_filehandle=self.log_fh)
self.scaff_graph_ok = self._parse_bam(self.sequences,
self.final_assembly_bam,
self.min_scaff_depth,
self.max_insert)
print('Scaffolding graph is OK:',
self.scaff_graph_ok,
file=self.log_fh)
if self.clean:
for suffix in ['soft_clipped', 'unmapped_mates', 'scaff']:
filename = self.final_assembly_bam + '.' + suffix
print('Deleting file', filename, file=self.log_fh)
os.unlink(filename)
# This is to make this object picklable, to keep multithreading happy
self.log_fh = None