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["-cluster"] = 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 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 get_aligned_muscle(seq1,seq2):
"""Returns aligned sequences and frac_same using MUSCLE.
THis needs to be moved to the muscle app controller
"""
outname = get_tmp_filename()
res = muscle_seqs([seq1,seq2], add_seq_names=True, WorkingDir="/tmp", out_filename=outname)
#raise ValueError, res['StdErr'].read()
#raise ValueError, res
seq1_aligned,seq2_aligned =list(MinimalFastaParser(res['MuscleOut'].read()))
res.cleanUp()
del(res)
seq1_aligned = seq1_aligned[1][1:]
seq2_aligned = seq2_aligned[1][1:]
frac_same = sum(array(seq1_aligned) == array(seq2_aligned))/min(len(seq1),
len(seq2))
return seq1_aligned,seq2_aligned,frac_same
