def test_launch_workflow_incorrect_trim(self):
"""
test if we get the warning when trim length
is too long
"""
# index the 70% rep. set database
ref_fp = join(self.test_data_dir, '70_otus.fasta')
ref_db_fp = build_index_sortmerna(ref_fp=(ref_fp, ),
working_dir=self.working_dir)
seqs_fp = self.seqs_s1_fp
output_fp = self.working_dir
mean_error = 0.005
error_dist = get_default_error_profile()
indel_prob = 0.01
indel_max = 3
min_size = 2
# trim length longer than sequences
trim_length = 151
left_trim_length = 0
threads = 1
with self.assertWarns(UserWarning):
launch_workflow(seqs_fp=seqs_fp,
working_dir=output_fp,
mean_error=mean_error,
error_dist=error_dist,
indel_prob=indel_prob,
indel_max=indel_max,
trim_length=trim_length,
left_trim_length=left_trim_length,
min_size=min_size,
ref_fp=(ref_fp, ),
ref_db_fp=ref_db_fp,
threads_per_sample=threads)
def test_launch_workflow_skip_trim(self):
# index the 70% rep. set database
ref_fp = join(self.test_data_dir, '70_otus.fasta')
ref_db_fp = build_index_sortmerna(ref_fp=(ref_fp, ),
working_dir=self.working_dir)
seqs_fp = self.seqs_s1_fp
output_fp = self.working_dir
mean_error = 0.005
error_dist = get_default_error_profile()
indel_prob = 0.01
indel_max = 3
min_size = 2
# trim length longer than sequences
trim_length = -1
left_trim_length = 0
threads = 1
output_fp = launch_workflow(seqs_fp=seqs_fp,
working_dir=output_fp,
mean_error=mean_error,
error_dist=error_dist,
indel_prob=indel_prob,
indel_max=indel_max,
trim_length=trim_length,
left_trim_length=left_trim_length,
min_size=min_size,
ref_fp=(ref_fp, ),
ref_db_fp=ref_db_fp,
threads_per_sample=threads)
exp = Sequence.read(self.no_trim_res, format='fasta')
res = Sequence.read(output_fp, format='fasta')
self.assertEqual(exp, res)
def run_workflow_try(self, simfilename, origfilename,
ref_fp, ref_db_fp, threads=1,
trim_length=100):
"""Test launching the complete workflow using simulated sequences
and compare to original ground truth.
Parameters
----------
simfilename : str
name of the simulated reads fasta file
origfilename : str
name of the fasta file with the ground truth sequences
ref_fp : list of str
list of the reference database files
def_db_fp : list of str
list of the indexed database files or None to create them
threads : int
number of threads to use (default=1)
trim_length : int
length of sequences to trim to (default=100)
"""
seqs_fp = simfilename
output_fp = self.working_dir
mean_error = 0.005
error_dist = get_default_error_profile()
indel_prob = 0.01
indel_max = 3
min_size = 2
negate = False
nochimera = launch_workflow(seqs_fp=seqs_fp, working_dir=output_fp,
mean_error=mean_error,
error_dist=error_dist,
indel_prob=indel_prob,
indel_max=indel_max,
trim_length=trim_length,
min_size=min_size,
ref_fp=(ref_fp,),
ref_db_fp=ref_db_fp,
negate=negate,
threads_per_sample=threads)
# get the trimmed ground truth sequences
orig_seqs = [item[1] for item in sequence_generator(origfilename)]
orig_seqs = [item[:trim_length].upper() for item in orig_seqs]
output_filename = 'final.biom'
output_table_fp = join(output_fp, output_filename)
create_otu_table(output_table_fp, [(nochimera, seqs_fp)])
table_obs = load_table(output_table_fp)
outseqs = table_obs.ids(axis='observation')
outseqs = list(outseqs)
outseqs.sort()
orig_seqs.sort()
# test we see all ground truth sequences and no other
self.assertEqual(outseqs, orig_seqs)
def run_workflow_try(self, simfilename, origfilename, ref_fp, ref_db_fp):
"""Test launching the complete workflow using simulated sequences
and compare to original ground truth.
Parameters
----------
simfilename : str
name of the simulated reads fasta file
origfilename : str
name of the fasta file with the ground truth sequences
"""
seqs_fp = simfilename
output_fp = self.working_dir
read_error = 0.05
mean_error = 0.005
error_dist = None
indel_prob = 0.01
indel_max = 3
trim_length = 100
min_size = 2
negate = False
threads = 1
delim = '_'
biom_fp = launch_workflow(seqs_fp, output_fp, read_error, mean_error,
error_dist, indel_prob, indel_max,
trim_length, min_size, (ref_fp, ), ref_db_fp,
negate, threads, delim)
# get the trimmed ground truth sequences
with open(origfilename, 'U') as f:
orig_seqs = [item[1] for item in parse_fasta(f)]
orig_seqs = [item[:trim_length].upper() for item in orig_seqs]
table_obs = load_table(biom_fp)
outseqs = table_obs.ids(axis='observation')
# test we see all ground truth sequences and no other
self.assertItemsEqual(outseqs, orig_seqs)
