def test_assemble_with_spades_fail(self):
'''test _assemble_with_spades handles spades fail'''
reads1 = os.path.join(
data_dir, 'assembly_test_assemble_with_spades_fails_reads_1.fq')
reads2 = os.path.join(
data_dir, 'assembly_test_assemble_with_spades_fails_reads_2.fq')
tmp_dir = 'tmp.test_assemble_with_spades_fail'
tmp_log = 'tmp.test_assemble_with_spades_fail.log'
with open(tmp_log, 'w') as tmp_log_fh:
print('First line', file=tmp_log_fh)
shutil.rmtree(tmp_dir, ignore_errors=True)
a = assembly.Assembly(reads1,
reads2,
'not needed',
'not needed',
tmp_dir,
'not_needed_for_this_test.fa',
'not_needed_for_this_test.bam',
tmp_log_fh,
'not needed',
assembler="spades",
spades_options=" --only-assembler")
a._assemble_with_spades()
self.assertFalse(a.assembled_ok)
shutil.rmtree(tmp_dir, ignore_errors=True)
os.unlink(tmp_log)
def test_assemble_with_fermilite(self):
'''test _assemble_with_fermilite'''
reads1 = os.path.join(data_dir,
'assembly_assemble_with_fermilite.reads_1.fq')
reads2 = os.path.join(data_dir,
'assembly_assemble_with_fermilite.reads_2.fq')
expected_log = os.path.join(
data_dir, 'assembly_assemble_with_fermilite.expected.log')
expected_fa = os.path.join(
data_dir, 'assembly_assemble_with_fermilite.expected.fa')
tmp_dir = 'tmp.test_assemble_with_fermilite'
tmp_log = 'tmp.test_assemble_with_fermilite.log'
tmp_log_fh = open(tmp_log, 'w')
print('First line', file=tmp_log_fh)
a = assembly.Assembly(reads1, reads2, 'not needed', 'not needed',
tmp_dir, 'not_needed_for_this_test.fa',
'not_needed_for_this_test.bam', tmp_log_fh,
'not needed')
a._assemble_with_fermilite()
self.assertTrue(a.assembled_ok)
tmp_log_fh.close()
self.assertTrue(filecmp.cmp(expected_log, tmp_log, shallow=False))
self.assertTrue(
filecmp.cmp(expected_fa,
os.path.join(tmp_dir, 'debug_all_contigs.fa'),
shallow=False))
shutil.rmtree(tmp_dir)
os.unlink(tmp_log)
def test_assemble_with_spades(self):
'''test _assemble_with_spades'''
reads1 = os.path.join(data_dir,
'assembly_test_assemble_with_spades_reads_1.fq')
reads2 = os.path.join(data_dir,
'assembly_test_assemble_with_spades_reads_2.fq')
tmp_dir = 'tmp.test_assemble_with_spades'
tmp_log = 'tmp.test_assemble_with_spades.log'
with open(tmp_log, 'w') as tmp_log_fh:
print('First line', file=tmp_log_fh)
shutil.rmtree(tmp_dir, ignore_errors=True)
#using spades_options=" --only-assembler" because error correction cannot determine quality offset on this
#artificial dataset
a = assembly.Assembly(reads1,
reads2,
'not needed',
'not needed',
tmp_dir,
'not_needed_for_this_test.fa',
'not_needed_for_this_test.bam',
tmp_log_fh,
'not needed',
assembler="spades",
spades_options=" --only-assembler")
a._assemble_with_spades()
self.assertTrue(a.assembled_ok)
shutil.rmtree(tmp_dir, ignore_errors=True)
os.unlink(tmp_log)
def _run(self):
print('{:_^79}'.format(' LOG FILE START ' + self.name + ' '),
file=self.log_fh,
flush=True)
if self.total_reads == 0:
print('No reads left after filtering with cdhit',
file=self.log_fh,
flush=True)
self.assembled_ok = False
else:
wanted_reads = self._number_of_reads_for_assembly(
self.longest_ref_length, self.reads_insert,
self.total_reads_bases, self.total_reads,
self.assembly_coverage)
made_reads = self._make_reads_for_assembly(
wanted_reads,
self.total_reads,
self.all_reads1,
self.all_reads2,
self.reads_for_assembly1,
self.reads_for_assembly2,
random_seed=self.random_seed)
print('\nUsing',
made_reads,
'from a total of',
self.total_reads,
'for assembly.',
file=self.log_fh,
flush=True)
print('Assembling reads:', file=self.log_fh, flush=True)
self._update_threads()
self.assembly = assembly.Assembly(
self.reads_for_assembly1,
self.reads_for_assembly2,
self.reference_fa,
self.references_fa,
self.assembly_dir,
self.final_assembly_fa,
self.final_assembly_bam,
self.log_fh,
self.all_refs_fasta,
contig_name_prefix=self.name,
assembler=self.assembler,
extern_progs=self.extern_progs,
clean=self.clean,
spades_mode=self.spades_mode,
spades_options=self.spades_options,
threads=self.threads)
self.assembly.run()
self.assembled_ok = self.assembly.assembled_ok
self._clean_file(self.reads_for_assembly1)
self._clean_file(self.reads_for_assembly2)
if self.clean:
print('Deleting Assembly directory',
self.assembly_dir,
file=self.log_fh,
flush=True)
shutil.rmtree(self.assembly_dir, ignore_errors=True)
if self.assembled_ok and self.assembly.ref_seq_name is not None:
self.ref_sequence = self.refdata.sequence(
self.assembly.ref_seq_name)
is_gene, is_variant_only = self.refdata.sequence_type(
self.ref_sequence.id)
self.is_gene = '1' if is_gene == 'p' else '0'
self.is_variant_only = '1' if is_variant_only else '0'
print('\nAssembly was successful\n\nMapping reads to assembly:',
file=self.log_fh,
flush=True)
self._update_threads()
mapping.run_bowtie2(
self.all_reads1,
self.all_reads2,
self.final_assembly_fa,
self.final_assembly_bam[:-4],
threads=self.threads,
sort=True,
bowtie2=self.extern_progs.exe('bowtie2'),
bowtie2_preset='very-sensitive-local',
bowtie2_version=self.extern_progs.version('bowtie2'),
verbose=True,
verbose_filehandle=self.log_fh)
if self.assembly.has_contigs_on_both_strands:
self.status_flag.add('hit_both_strands')
print('\nMaking and checking scaffold graph',
file=self.log_fh,
flush=True)
if not self.assembly.scaff_graph_ok:
self.status_flag.add('scaffold_graph_bad')
print('Comparing assembly against reference sequence',
file=self.log_fh,
flush=True)
self.assembly_compare = assembly_compare.AssemblyCompare(
self.final_assembly_fa,
self.assembly.sequences,
self.reference_fa,
self.ref_sequence,
self.assembly_compare_prefix,
self.refdata,
nucmer_min_id=self.nucmer_min_id,
nucmer_min_len=self.nucmer_min_len,
nucmer_breaklen=self.nucmer_breaklen,
assembled_threshold=self.assembled_threshold,
unique_threshold=self.unique_threshold,
max_gene_nt_extend=self.max_gene_nt_extend,
)
self.assembly_compare.run()
self.status_flag = self.assembly_compare.update_flag(
self.status_flag)
allowed_ctg_pos, allowed_ref_pos = assembly_compare.AssemblyCompare.nucmer_hits_to_ref_and_qry_coords(
self.assembly_compare.nucmer_hits)
assembly_variants_obj = assembly_variants.AssemblyVariants(
self.refdata, self.assembly_compare.nucmer_snps_file)
self.assembly_variants = assembly_variants_obj.get_variants(
self.ref_sequence.id, allowed_ctg_pos, allowed_ref_pos)
for var_list in self.assembly_variants.values():
for var in var_list:
if var[3] not in ['.', 'SYN', None]:
self.status_flag.add('has_variant')
break
if self.status_flag.has('has_variant'):
break
print('\nCalling variants with samtools:',
file=self.log_fh,
flush=True)
self.samtools_vars = samtools_variants.SamtoolsVariants(
self.final_assembly_fa,
self.final_assembly_bam,
self.samtools_vars_prefix,
log_fh=self.log_fh,
min_var_read_depth=self.min_var_read_depth,
min_second_var_read_depth=self.min_second_var_read_depth,
max_allele_freq=self.max_allele_freq)
self.samtools_vars.run()
self.total_contig_depths = self.samtools_vars.total_depth_per_contig(
self.samtools_vars.contig_depths_file)
self.variants_from_samtools = self.samtools_vars.variants_in_coords(
self.assembly_compare.assembly_match_coords(),
self.samtools_vars.vcf_file)
if len(self.variants_from_samtools):
self.status_flag.add('variants_suggest_collapsed_repeat')
elif not self.assembled_ok:
print('\nAssembly failed\n', file=self.log_fh, flush=True)
self.status_flag.add('assembly_fail')
elif self.assembly.ref_seq_name is None:
print('\nCould not get closest reference sequence\n',
file=self.log_fh,
flush=True)
self.status_flag.add('ref_seq_choose_fail')
try:
self.report_lines = report.report_lines(self)
except:
print('Error making report for cluster ',
self.name,
'... traceback:',
file=sys.stderr)
traceback.print_exc(file=sys.stderr)
raise Error('Error making report for cluster ' + self.name)
self._clean()
atexit.unregister(self._atexit)