def setUp(self):
# create a list of files to cleanup
self._paths_to_clean_up = []
self._dirs_to_clean_up = []
# load query seqs
self.seqs = Alignment(parse_fasta(QUERY_SEQS.split()))
# generate temp filename
tmp_dir='/tmp'
self.outfile = get_tmp_filename(tmp_dir)
# create and write out reference sequence file
self.outfasta=splitext(self.outfile)[0]+'.fasta'
fastaout=open(self.outfasta,'w')
fastaout.write(REF_SEQS)
fastaout.close()
self._paths_to_clean_up.append(self.outfasta)
# create and write out starting tree file
self.outtree=splitext(self.outfile)[0]+'.tree'
treeout=open(self.outtree,'w')
treeout.write(REF_TREE)
treeout.close()
self._paths_to_clean_up.append(self.outtree)
def cluster_seqs(seqs,
neighbor_join=False,
params={},
add_seq_names=True,
WorkingDir=None,
SuppressStderr=None,
SuppressStdout=None,
max_chars=1000000,
max_hours=1.0,
constructor=PhyloNode,
clean_up=True
):
"""Muscle cluster list of sequences.
seqs: either file name or list of sequence objects or list of strings or
single multiline string containing sequences.
Addl docs coming soon
"""
num_seqs = len(seqs)
if num_seqs < 2:
raise ValueError, "Muscle requres 2 or more sequences to cluster."
num_chars = sum(map(len, seqs))
if num_chars > max_chars:
params["-maxiters"] = 2
params["-diags1"] = True
params["-sv"] = True
#params["-distance1"] = "kmer6_6"
#params["-distance1"] = "kmer20_3"
#params["-distance1"] = "kbit20_3"
print "lots of chars, using fast align", num_chars
params["-maxhours"] = max_hours
#params["-maxiters"] = 10
#cluster_type = "upgmb"
#if neighbor_join:
# cluster_type = "neighborjoining"
params["-clusteronly"] = True
params["-tree1"] = get_tmp_filename(WorkingDir)
muscle_res = muscle_seqs(seqs,
params=params,
add_seq_names=add_seq_names,
WorkingDir=WorkingDir,
SuppressStderr=SuppressStderr,
SuppressStdout=SuppressStdout)
tree = DndParser(muscle_res["Tree1Out"], constructor=constructor)
if clean_up:
muscle_res.cleanUp()
return tree
def build_tree_from_alignment(aln, moltype, best_tree=False, params={}):
"""Returns a tree from Alignment object aln.
aln: an xxx.Alignment object, or data that can be used to build one.
moltype: cogent.core.moltype.MolType object
best_tree: best_tree suppport is currently not implemented
params: dict of parameters to pass in to the RAxML app controller.
The result will be an xxx.Alignment object, or None if tree fails.
"""
if best_tree:
raise NotImplementedError
if '-m' not in params:
if moltype == DNA or moltype == RNA:
#params["-m"] = 'GTRMIX'
# in version 7.2.3, GTRMIX is no longer supported but says GTRCAT
# behaves like GTRMIX (http://www.phylo.org/tools/raxmlhpc2.html)
params["-m"] = 'GTRGAMMA'
elif moltype == PROTEIN:
params["-m"] = 'PROTGAMMAmatrixName'
else:
raise ValueError, "Moltype must be either DNA, RNA, or PROTEIN"
if not hasattr(aln, 'toPhylip'):
aln = Alignment(aln)
seqs, align_map = aln.toPhylip()
# generate temp filename for output
params["-w"] = "/tmp/"
params["-n"] = get_tmp_filename().split("/")[-1]
params["-k"] = True
params["-p"] = randint(1,100000)
params["-x"] = randint(1,100000)
ih = '_input_as_multiline_string'
raxml_app = Raxml(params=params,
InputHandler=ih,
WorkingDir=None,
SuppressStderr=True,
SuppressStdout=True)
raxml_result = raxml_app(seqs)
tree = DndParser(raxml_result['Bootstrap'], constructor=PhyloNode)
for node in tree.tips():
node.Name = align_map[node.Name]
raxml_result.cleanUp()
return tree
def aln_tree_seqs(seqs,
input_handler=None,
tree_type='neighborjoining',
params={},
add_seq_names=True,
WorkingDir=None,
SuppressStderr=None,
SuppressStdout=None,
max_hours=5.0,
constructor=PhyloNode,
clean_up=True
):
"""Muscle align sequences and report tree from iteration2.
Unlike cluster_seqs, returns tree2 which is the tree made during the
second muscle iteration (it should be more accurate that the cluster from
the first iteration which is made fast based on k-mer words)
seqs: either file name or list of sequence objects or list of strings or
single multiline string containing sequences.
tree_type: can be either neighborjoining (default) or upgmb for UPGMA
clean_up: When true, will clean up output files
"""
params["-maxhours"] = max_hours
if tree_type:
params["-cluster2"] = tree_type
params["-tree2"] = get_tmp_filename(WorkingDir)
params["-out"] = get_tmp_filename(WorkingDir)
muscle_res = muscle_seqs(seqs,
input_handler=input_handler,
params=params,
add_seq_names=add_seq_names,
WorkingDir=WorkingDir,
SuppressStderr=SuppressStderr,
SuppressStdout=SuppressStdout)
tree = DndParser(muscle_res["Tree2Out"], constructor=constructor)
aln = [line for line in muscle_res["MuscleOut"]]
if clean_up:
muscle_res.cleanUp()
return tree, aln
def raxml_alignment(align_obj,
raxml_model="GTRCAT",
params={},
SuppressStderr=True,
SuppressStdout=True):
"""Run raxml on alignment object
align_obj: Alignment object
params: you can set any params except -w and -n
returns: tuple (phylonode,
parsimonyphylonode,
log likelihood,
total exec time)
"""
# generate temp filename for output
params["-w"] = "/tmp/"
params["-n"] = get_tmp_filename().split("/")[-1]
params["-m"] = raxml_model
params["-p"] = randint(1,100000)
ih = '_input_as_multiline_string'
seqs, align_map = align_obj.toPhylip()
#print params["-n"]
# set up command
raxml_app = Raxml(
params=params,
InputHandler=ih,
WorkingDir=None,
SuppressStderr=SuppressStderr,
SuppressStdout=SuppressStdout)
# run raxml
ra = raxml_app(seqs)
# generate tree
tree_node = DndParser(ra["Result"])
# generate parsimony tree
parsimony_tree_node = DndParser(ra["ParsimonyTree"])
# extract log likelihood from log file
log_file = ra["Log"]
total_exec_time = exec_time = log_likelihood = 0.0
for line in log_file:
exec_time, log_likelihood = map(float, line.split())
total_exec_time += exec_time
# remove output files
ra.cleanUp()
return tree_node, parsimony_tree_node, log_likelihood, total_exec_time
def test_insert_sequences_into_tree(self):
"""Inserts sequences into Tree using params - test handles tree-insertion"""
# generate temp filename for output
outfname=splitext(get_tmp_filename('/tmp/'))[0]
# create starting tree
outtreefname=outfname+'.tre'
outtree=open(outtreefname,'w')
outtree.write(REF_TREE)
outtree.close()
# set params for tree-insertion
params={}
params["-w"]="/tmp/"
params["-n"] = get_tmp_filename().split("/")[-1]
params["-f"] = 'v'
#params["-G"] = '0.25'
params["-t"] = outtreefname
params["-m"] = 'GTRGAMMA'
aln_ref_query=get_align_for_phylip(StringIO(PHYLIP_FILE_DNA_REF_QUERY))
aln = Alignment(aln_ref_query)
seqs, align_map = aln.toPhylip()
tree = insert_sequences_into_tree(seqs, DNA, params=params,
write_log=False)
for node in tree.tips():
removed_query_str=re.sub('QUERY___','',str(node.Name))
new_node_name=re.sub('___\d+','',str(removed_query_str))
if new_node_name in align_map:
node.Name = align_map[new_node_name]
self.assertTrue(isinstance(tree, PhyloNode))
self.assertEqual(tree.getNewick(with_distances=True),RESULT_TREE)
self.assertEqual(len(tree.tips()), 7)
self.assertRaises(NotImplementedError, build_tree_from_alignment, \
self.align1, RNA, True)
remove(outtreefname)
def insert_sequences_into_tree(seqs, moltype, params={},
write_log=True):
"""Insert sequences into Tree.
aln: an xxx.Alignment object, or data that can be used to build one.
moltype: cogent.core.moltype.MolType object
params: dict of parameters to pass in to the RAxML app controller.
The result will be a tree.
"""
ih = '_input_as_multiline_string'
raxml_app = Raxml(params=params,
InputHandler=ih,
WorkingDir=None,
SuppressStderr=False,
SuppressStdout=False,
HALT_EXEC=False)
raxml_result = raxml_app(seqs)
# write a log file
if write_log:
log_fp = join(params["-w"],'log_raxml_'+split(get_tmp_filename())[-1])
log_file=open(log_fp,'w')
log_file.write(raxml_result['StdOut'].read())
log_file.close()
'''
# getting setup since parsimony doesn't output tree..only jplace, however
# it is currently corrupt
# use guppy to convert json file into a placement tree
guppy_params={'tog':None}
new_tree=build_tree_from_json_using_params(raxml_result['json'].name, \
output_dir=params["-w"], \
params=guppy_params)
'''
# get tree from 'Result Names'
new_tree=raxml_result['Result'].readlines()
filtered_tree=re.sub('\[I\d+\]','',str(new_tree))
tree = DndParser(filtered_tree, constructor=PhyloNode)
raxml_result.cleanUp()
return tree
def build_tree_from_distance_matrix(matrix, best_tree=False, params={}, working_dir="/tmp"):
"""Returns a tree from a distance matrix.
matrix: a square Dict2D object (cogent.util.dict2d)
best_tree: if True (default:False), uses a slower but more accurate
algorithm to build the tree.
params: dict of parameters to pass in to the Clearcut app controller.
The result will be an cogent.core.tree.PhyloNode object, or None if tree
fails.
"""
params["--out"] = get_tmp_filename(working_dir)
# Create instance of app controller, enable tree, disable alignment
app = Clearcut(
InputHandler="_input_as_multiline_string",
params=params,
WorkingDir=working_dir,
SuppressStdout=True,
SuppressStderr=True,
)
# Turn off input as alignment
app.Parameters["-a"].off()
# Input is a distance matrix
app.Parameters["-d"].on()
if best_tree:
app.Parameters["-N"].on()
# Turn the dict2d object into the expected input format
matrix_input, int_keys = _matrix_input_from_dict2d(matrix)
# Collect result
result = app(matrix_input)
# Build tree
tree = DndParser(result["Tree"].read(), constructor=PhyloNode)
# reassign to original names
for node in tree.tips():
node.Name = int_keys[node.Name]
# Clean up
result.cleanUp()
del (app, result, params)
return tree
def insert_sequences_into_tree(aln, moltype, params={},
write_log=True):
"""Returns a tree from Alignment object aln.
aln: an xxx.Alignment object, or data that can be used to build one.
moltype: cogent.core.moltype.MolType object
params: dict of parameters to pass in to the RAxML app controller.
The result will be an xxx.Alignment object, or None if tree fails.
"""
# convert aln to phy since seq_names need fixed to run through pplacer
new_aln=get_align_for_phylip(StringIO(aln))
# convert aln to fasta in case it is not already a fasta file
aln2 = Alignment(new_aln)
seqs = aln2.toFasta()
ih = '_input_as_multiline_string'
pplacer_app = Pplacer(params=params,
InputHandler=ih,
WorkingDir=None,
SuppressStderr=False,
SuppressStdout=False)
pplacer_result = pplacer_app(seqs)
# write a log file
if write_log:
log_fp = join(params["--out-dir"],'log_pplacer_' + \
split(get_tmp_filename())[-1])
log_file=open(log_fp,'w')
log_file.write(pplacer_result['StdOut'].read())
log_file.close()
# use guppy to convert json file into a placement tree
guppy_params={'tog':None}
new_tree=build_tree_from_json_using_params(pplacer_result['json'].name, \
output_dir=params['--out-dir'], \
params=guppy_params)
pplacer_result.cleanUp()
return new_tree
def build_tree_from_alignment(aln, moltype=DNA, best_tree=False, params=None):
"""Returns a tree from Alignment object aln.
aln: a cogent.core.alignment.Alignment object, or data that can be used
to build one.
moltype: cogent.core.moltype.MolType object
best_tree: unsupported
params: dict of parameters to pass in to the Muscle app controller.
The result will be an cogent.core.tree.PhyloNode object, or None if tree
fails.
"""
# Create instance of app controller, enable tree, disable alignment
app = Muscle(InputHandler='_input_as_multiline_string', params=params, \
WorkingDir='/tmp')
app.Parameters['-clusteronly'].on()
app.Parameters['-tree1'].on(get_tmp_filename(app.WorkingDir))
app.Parameters['-seqtype'].on(moltype.label)
seq_collection = SequenceCollection(aln, MolType=moltype)
#Create mapping between abbreviated IDs and full IDs
int_map, int_keys = seq_collection.getIntMap()
#Create SequenceCollection from int_map.
int_map = SequenceCollection(int_map,MolType=moltype)
# Collect result
result = app(int_map.toFasta())
# Build tree
tree = DndParser(result['Tree1Out'].read(), constructor=PhyloNode)
for tip in tree.tips():
tip.Name = int_keys[tip.Name]
# Clean up
result.cleanUp()
del(seq_collection, app, result)
return tree
def align_unaligned_seqs(seqs, moltype=DNA, params=None):
"""Returns an Alignment object from seqs.
seqs: SequenceCollection object, or data that can be used to build one.
moltype: a MolType object. DNA, RNA, or PROTEIN.
params: dict of parameters to pass in to the Muscle app controller.
Result will be an Alignment object.
"""
if not params:
params = {}
#create SequenceCollection object from seqs
seq_collection = SequenceCollection(seqs,MolType=moltype)
#Create mapping between abbreviated IDs and full IDs
int_map, int_keys = seq_collection.getIntMap()
#Create SequenceCollection from int_map.
int_map = SequenceCollection(int_map,MolType=moltype)
#get temporary filename
params.update({'-out':get_tmp_filename()})
#Create Muscle app.
app = Muscle(InputHandler='_input_as_multiline_string',\
params=params)
#Get results using int_map as input to app
res = app(int_map.toFasta())
#Get alignment as dict out of results
alignment = dict(parse_fasta(res['MuscleOut']))
#Make new dict mapping original IDs
new_alignment = {}
for k,v in alignment.items():
new_alignment[int_keys[k]]=v
#Create an Alignment object from alignment dict
new_alignment = Alignment(new_alignment,MolType=moltype)
#Clean up
res.cleanUp()
del(seq_collection,int_map,int_keys,app,res,alignment,params)
return new_alignment
def setUp(self):
'''setup the files for testing pplacer'''
# create a list of files to cleanup
self._paths_to_clean_up = []
self._dirs_to_clean_up = []
# get a tmp filename to use
basename=splitext(get_tmp_filename())[0]
# create and write out RAxML stats file
self.stats_fname=basename+'.stats'
stats_out=open(self.stats_fname,'w')
stats_out.write(RAXML_STATS)
stats_out.close()
self._paths_to_clean_up.append(self.stats_fname)
# create and write out reference sequence file
self.refseq_fname=basename+'_refseqs.fasta'
refseq_out=open(self.refseq_fname,'w')
refseq_out.write(REF_SEQS)
refseq_out.close()
self._paths_to_clean_up.append(self.refseq_fname)
# create and write out query sequence file
self.query_fname=basename+'_queryseqs.fasta'
query_out=open(self.query_fname,'w')
query_out.write(QUERY_SEQS)
query_out.close()
self._paths_to_clean_up.append(self.query_fname)
# create and write out starting tree file
self.tree_fname=basename+'.tre'
tree_out=open(self.tree_fname,'w')
tree_out.write(REF_TREE)
tree_out.close()
self._paths_to_clean_up.append(self.tree_fname)
def build_tree_from_alignment(aln, moltype=DNA, best_tree=False, params={}, working_dir="/tmp"):
"""Returns a tree from Alignment object aln.
aln: an cogent.core.alignment.Alignment object, or data that can be used
to build one.
- Clearcut only accepts aligned sequences. Alignment object used to
handle unaligned sequences.
moltype: a cogent.core.moltype object.
- NOTE: If moltype = RNA, we must convert to DNA since Clearcut v1.0.8
gives incorrect results if RNA is passed in. 'U' is treated as an
incorrect character and is excluded from distance calculations.
best_tree: if True (default:False), uses a slower but more accurate
algorithm to build the tree.
params: dict of parameters to pass in to the Clearcut app controller.
The result will be an cogent.core.tree.PhyloNode object, or None if tree
fails.
"""
params["--out"] = get_tmp_filename(working_dir)
# Create instance of app controller, enable tree, disable alignment
app = Clearcut(
InputHandler="_input_as_multiline_string",
params=params,
WorkingDir=working_dir,
SuppressStdout=True,
SuppressStderr=True,
)
# Input is an alignment
app.Parameters["-a"].on()
# Turn off input as distance matrix
app.Parameters["-d"].off()
# If moltype = RNA, we must convert to DNA.
if moltype == RNA:
moltype = DNA
if best_tree:
app.Parameters["-N"].on()
# Turn on correct moltype
moltype_string = moltype.label.upper()
app.Parameters[MOLTYPE_MAP[moltype_string]].on()
# Setup mapping. Clearcut clips identifiers. We will need to remap them.
# Clearcut only accepts aligned sequences. Let Alignment object handle
# unaligned sequences.
seq_aln = Alignment(aln, MolType=moltype)
# get int mapping
int_map, int_keys = seq_aln.getIntMap()
# create new Alignment object with int_map
int_map = Alignment(int_map)
# Collect result
result = app(int_map.toFasta())
# Build tree
tree = DndParser(result["Tree"].read(), constructor=PhyloNode)
for node in tree.tips():
node.Name = int_keys[node.Name]
# Clean up
result.cleanUp()
del (seq_aln, app, result, int_map, int_keys, params)
return tree
def align_two_alignments(aln1, aln2, params=None):
"""Returns an Alignment object from two existing Alignments.
aln1, aln2: cogent.core.alignment.Alignment objects, or data that can be
used to build them.
params: dict of parameters to pass in to the Muscle app controller.
"""
if not params:
params = {}
#create SequenceCollection object from aln1
aln1_collection = SequenceCollection(aln1)
#Create mapping between abbreviated IDs and full IDs
aln1_int_map, aln1_int_keys = aln1_collection.getIntMap(prefix='aln1_')
#Create SequenceCollection from int_map.
aln1_int_map = SequenceCollection(aln1_int_map)
#create SequenceCollection object from aln2
aln2_collection = SequenceCollection(aln2)
#Create mapping between abbreviated IDs and full IDs
aln2_int_map, aln2_int_keys = aln2_collection.getIntMap(prefix='aln2_')
#Create SequenceCollection from int_map.
aln2_int_map = SequenceCollection(aln2_int_map)
#set output and profile options
params.update({'-out':get_tmp_filename(), '-profile':True})
#save aln1 to tmp file
aln1_filename = get_tmp_filename()
aln1_out = open(aln1_filename,'w')
aln1_out.write(aln1_int_map.toFasta())
aln1_out.close()
#save aln2 to tmp file
aln2_filename = get_tmp_filename()
aln2_out = open(aln2_filename, 'w')
aln2_out.write(aln2_int_map.toFasta())
aln2_out.close()
#Create Muscle app and get results
app = Muscle(InputHandler='_input_as_multifile', params=params)
res = app((aln1_filename, aln2_filename))
#Get alignment as dict out of results
alignment = dict(parse_fasta(res['MuscleOut']))
#Make new dict mapping original IDs
new_alignment = {}
for k,v in alignment.items():
if k in aln1_int_keys:
new_alignment[aln1_int_keys[k]] = v
else:
new_alignment[aln2_int_keys[k]] = v
#Create an Alignment object from alignment dict
new_alignment = Alignment(new_alignment)
#Clean up
res.cleanUp()
del(aln1_collection, aln1_int_map, aln1_int_keys)
del(aln2_collection, aln2_int_map, aln2_int_keys)
del(app, res, alignment, params)
remove(aln1_filename)
remove(aln2_filename)
return new_alignment
def assign_taxonomy(dataPath, reference_sequences_fp, id_to_taxonomy_fp, read_1_seqs_fp, read_2_seqs_fp, single_ok=False, no_single_ok_generic=False,
header_id_regex=None, read_id_regex = "\S+\s+(\S+)", amplicon_id_regex = "(\S+)\s+(\S+?)\/",
output_fp=None, log_path=None, HALT_EXEC=False, base_tmp_dir = '/tmp'):
"""Assign taxonomy to each sequence in data with the RTAX classifier
# data: open fasta file object or list of fasta lines
dataPath: path to a fasta file
output_fp: path to write output; if not provided, result will be
returned in a dict of {seq_id:(taxonomy_assignment,confidence)}
"""
usearch_command = "usearch"
if not (exists(usearch_command) or app_path(usearch_command)):
raise ApplicationNotFoundError,\
"Cannot find %s. Is it installed? Is it in your path?"\
% usearch_command
my_tmp_dir = get_tmp_filename(tmp_dir=base_tmp_dir,prefix='rtax_',suffix='',result_constructor=str)
os.makedirs(my_tmp_dir)
try:
# RTAX classifier doesn't necessarily preserve identifiers
# it reports back only the id extracted as $1 using header_id_regex
# since rtax takes the original unclustered sequence files as input,
# the usual case is that the regex extracts the amplicon ID from the second field
# Use lookup table
read_1_id_to_orig_id = {}
readIdExtractor = re.compile(read_id_regex) # OTU clustering produces ">clusterID read_1_id"
data = open(dataPath,'r')
for seq_id, seq in parse_fasta(data):
# apply the regex
extract = readIdExtractor.match(seq_id)
if extract is None:
stderr.write("Matched no ID with read_id_regex " + read_id_regex +" in '" + seq_id + "' from file " + dataPath + "\n")
else:
read_1_id_to_orig_id[extract.group(1)] = seq_id
#stderr.write(extract.group(1) + " => " + seq_id + "\n")
#seq_id_lookup[seq_id.split()[1]] = seq_id
data.close()
# make list of amplicon IDs to pass to RTAX
id_list_fp = open(my_tmp_dir+"/ampliconIdsToClassify", "w")
# Establish mapping of amplicon IDs to read_1 IDs
# simultaneously write the amplicon ID file for those IDs found in the input mapping above
amplicon_to_read_1_id = {}
ampliconIdExtractor = re.compile(amplicon_id_regex) # split_libraries produces >read_1_id ampliconID/1 ... // see also assign_taxonomy 631
read_1_data = open(read_1_seqs_fp,'r')
for seq_id, seq in parse_fasta(read_1_data):
# apply the regex
extract = ampliconIdExtractor.match(seq_id)
if extract is None:
stderr.write("Matched no ID with amplicon_id_regex " + amplicon_id_regex + " in '" + seq_id + "' from file " + read_1_seqs_fp + "\n")
else:
read_1_id = extract.group(1)
amplicon_id = extract.group(2)
try:
amplicon_to_read_1_id[amplicon_id] = read_1_id
bogus = read_1_id_to_orig_id[read_1_id] # verify that the id is valid
id_list_fp.write('%s\n' % (amplicon_id))
except KeyError:
pass
data.close()
id_list_fp.close()
app = Rtax(HALT_EXEC=HALT_EXEC)
temp_output_file = tempfile.NamedTemporaryFile(
prefix='RtaxAssignments_', suffix='.txt')
app.Parameters['-o'].on(temp_output_file.name)
app.Parameters['-r'].on(reference_sequences_fp)
app.Parameters['-t'].on(id_to_taxonomy_fp)
# app.Parameters['-d'].on(delimiter)
app.Parameters['-l'].on(id_list_fp.name) # these are amplicon IDs
app.Parameters['-a'].on(read_1_seqs_fp)
if read_2_seqs_fp is not None:
app.Parameters['-b'].on(read_2_seqs_fp)
app.Parameters['-i'].on(header_id_regex)
app.Parameters['-m'].on(my_tmp_dir)
if single_ok: app.Parameters['-f'].on();
if no_single_ok_generic: app.Parameters['-g'].on();
#app.Parameters['-v'].on()
app_result = app()
if log_path:
f=open(log_path, 'a')
errString=''.join(app_result['StdErr'].readlines()) + '\n'
f.write(errString)
f.close()
assignments = {}
# restore original sequence IDs with spaces
for line in app_result['Assignments']:
toks = line.strip().split('\t')
rtax_id = toks.pop(0)
if len(toks):
bestpcid = toks.pop(0) # ignored
lineage = toks
# RTAX does not provide a measure of confidence. We could pass one in,
# based on the choice of primers, or even look it up on the fly in the tables
# from the "optimal primers" paper; but it would be the same for every
# query sequence anyway.
# we could also return bestpcid, but that's not the same thing as confidence.
confidence = 1.0
read_1_id = amplicon_to_read_1_id[rtax_id]
orig_id = read_1_id_to_orig_id[read_1_id]
if lineage:
assignments[orig_id] = (';'.join(lineage), confidence)
else:
assignments[orig_id] = ('Unclassified', 1.0)
if output_fp:
try:
output_file = open(output_fp, 'w')
except OSError:
raise OSError("Can't open output file for writing: %s" % output_fp)
for seq_id, assignment in assignments.items():
lineage, confidence = assignment
output_file.write(
'%s\t%s\t%1.3f\n' % (seq_id, lineage, confidence))
output_file.close()
return None
else:
return assignments
finally:
try:
rmtree(my_tmp_dir)
except OSError:
pass