def mafft(fasta_file, output_dir, output_ext, seq_type, cpus, anysymbol):
"""Align sequences."""
in_path = fasta_file
if seq_type == 'aa':
in_path = bio.adjust_aa_seqs(fasta_file, output_dir)
cmd = [
'mafft', '--amino' if seq_type == 'aa' else '--nuc',
'--thread {}'.format(cpus), '--anysymbol' if anysymbol else ''
]
if (bio.fasta_record_count(in_path) >= bio.SEQ_COUNT_CUTOFF
or bio.longest_fasta_seq(in_path) >= bio.SEQ_LEN_CUTOFF):
cmd.append('--auto')
else:
cmd += [
'--genafpair', '--maxiterate {}'.format(MAX_ITERATE),
'--anysymbol' if anysymbol else ''
]
cmd.append(in_path)
cmd = ' '.join(cmd)
aligned = util.file_name(fasta_file, output_ext)
with util.cd(output_dir):
result = subprocess.check_output(cmd, shell=True)
with open(aligned, 'wb') as out_file:
out_file.write(result)
return aligned
def treeshrink(tree_file, output_dir, output_ext, quantiles):
"""Remove long branches from a tree."""
subdir = util.file_name(tree_file)
cmd = ' '.join([
'run_treeshrink.py',
'--tree {}'.format(tree_file),
'--centroid',
'--mode per-gene',
'--quantiles {}'.format(quantiles),
'--outdir {}'.format(subdir),
'--tempdir {}'.format(subdir)])
with util.cd(output_dir):
subprocess.check_call(cmd, shell=True)
mask = util.file_name(subdir + '_*', ext=EXT_IN, dir_=subdir)
tree_src = glob(mask)[0]
tree_dst = util.file_name(tree_file, output_ext + EXT_OUT)
with open(tree_src) as in_file, open(tree_dst, 'w') as out_file:
content = in_file.read()
out_file.write(content.replace("'", ''))
rmtree(subdir)
return tree_dst
def raxml_ng_bs(fasta_file,
output_dir,
temp_dir,
seq_type,
cpus,
seed,
output_ext,
replicates=100):
"""Build a bootstrapped tree with raxml."""
model = "Blosum62" if seq_type == "aa" else "GTR"
tree = util.file_name(fasta_file, output_ext, output_dir)
cmd = ' '.join([
'raxml-ng', '-T {}'.format(cpus), '-f a', '-x {}'.format(seed),
'-p {}'.format(seed), '-m {}'.format(model),
'-# {}'.format(replicates), '-s {}'.format(fasta_file),
'-n {}'.format(tree)
])
with util.cd(temp_dir):
subprocess.check_call(cmd, shell=True)
tree_src = join('RAxML_bipartitions.' + tree)
tree_dst = join(output_dir, tree)
move(tree_src, tree_dst)
return tree_dst
def mask_tips(tree_file, output_dir, output_ext):
"""Wrap tree tip removal."""
tree = Phylo.read(tree_file, 'newick')
mask_monophyletic_tips(tree)
output = util.file_name(tree_file, output_ext)
with util.cd(output_dir):
Phylo.write(tree, output, 'newick')
return output
def pxrr(tree_file, output_dir):
"""Unroot the tree returned by treeshrink."""
unrooted = util.file_name(tree_file)
cmd = ' '.join([
'pxrr', '--unroot', '--treef {}'.format(tree_file),
'--outf {}'.format(unrooted)
])
with util.cd(output_dir):
subprocess.check_call(cmd, shell=True)
util.remove_files('phyx.logfile')
return unrooted
def cut_branches(tree_file, output_dir, output_ext, branch_cutoff, min_taxa):
"""Cut long internal branches."""
tree = Phylo.read(tree_file, 'newick')
subtrees = cut_deep(tree, branch_cutoff, min_taxa)
with util.cd(output_dir):
for i, subtree in enumerate(subtrees):
output = '{}_{}'.format(tree_file, i)
output = util.file_name(output, output_ext)
Phylo.write(subtree, output, 'newick')
return output
def fasttree(fasta_file, output_dir, output_ext, seq_type):
"""Build a tree with fasttree."""
cmd = ['fasttree', '-quiet']
cmd += ['-wag'] if seq_type == 'aa' else ['-nt', '-gtr']
cmd.append(fasta_file)
cmd = ' '.join(cmd)
tree_file = util.file_name(fasta_file, output_ext)
with util.cd(output_dir):
result = subprocess.check_output(cmd, shell=True)
with open(tree_file, 'wb') as out_file:
out_file.write(result)
return tree_file
def raxml(fasta_file, output_dir, output_ext, seq_type, cpus, seed):
"""Build a tree with raxml."""
model = "PROTCATWAG" if seq_type == "aa" else "GTRCAT"
tree = util.file_name(fasta_file, output_ext)
cmd = ' '.join([
'raxml', '-T {}'.format(cpus), '-p {}'.format(seed),
'-m {}'.format(model), '-s {}'.format(fasta_file), '-n {}'.format(tree)
])
with util.cd(output_dir):
subprocess.check_call(cmd, shell=True)
tree_src = 'RAxML_bestTree.' + tree
move(tree_src, tree)
util.remove_files('RAxML_*')
return tree
def raxml_ng(fasta_file, output_dir, temp_dir, seq_type, cpus, seed,
output_ext):
"""Build a tree with raxml."""
model = "Blosum62" if seq_type == "aa" else "GTR"
tree = util.file_name(fasta_file, output_ext)
cmd = ' '.join([
'raxml-ng', '-T {}'.format(cpus), '-p {}'.format(seed),
'-m {}'.format(model), '-s {}'.format(fasta_file), '-n {}'.format(tree)
])
with util.cd(temp_dir):
subprocess.check_call(cmd, shell=True)
tree_src = join('RAxML_bestTree.' + tree)
tree_dst = join(output_dir, tree)
move(tree_src, tree_dst)
return tree_dst
def prank(fasta_file, output_dir, temp_dir, seq_type):
"""Align sequences."""
in_path = fasta_file
if seq_type == 'aa':
in_path = bio.adjust_aa_seqs(fasta_file, temp_dir)
aligned = util.file_name(fasta_file, 'ortho.aln')
cmd = [
'prank',
'-d {}'.format(in_path),
'-o {}'.format(aligned),
'-protein' if seq_type == 'aa' else '-DNA',
]
cmd = ' '.join(cmd)
with util.cd(temp_dir):
result = subprocess.check_output(cmd)
with open(aligned, 'wb') as out_file:
out_file.write(result)
return aligned
def pasta(fasta_file, output_dir, output_ext, seq_type, cpus):
"""Align sequences."""
in_path = fasta_file
if seq_type == 'aa':
in_path = bio.adjust_aa_seqs(fasta_file, output_dir)
cmd = ' '.join([
which('run_pasta.py'),
'--datatype {}'.format('Protein' if seq_type == 'aa' else 'DNA'),
'--num-cpus {}'.format(cpus), "--input '{}'".format(in_path),
"--output-directory '{}'".format(abspath(output_dir))
])
with util.cd(output_dir):
subprocess.check_call(cmd, shell=True)
base_name = splitext(basename(fasta_file))[0]
temp_aligned = 'pastajob.marker001.' + base_name + EXT
aligned = base_name + output_ext
move(temp_aligned, aligned)
util.remove_files('pastajob*')
return aligned
def pxclsq(fasta_file, output_dir, output_ext, seq_type, min_occupancy,
min_len):
"""Filter aligned sequences for occupancy and length."""
ext = output_ext + EXT_PXCLSQ
temp_cleaned = util.file_name(fasta_file, ext)
cmd = ' '.join([
'pxclsq', '--aminoacid' if seq_type == 'aa' else '',
'--prop {}'.format(min_occupancy), '--seqf {}'.format(fasta_file),
'--outf {}'.format(basename(temp_cleaned))
])
cleaned = util.file_name(fasta_file, output_ext)
with util.cd(output_dir):
subprocess.check_call(cmd, shell=True)
with open(temp_cleaned) as in_file, open(cleaned, 'w') as out_file:
for header, seq in SimpleFastaParser(in_file):
if len(seq.replace('-', '')) >= min_len:
bio.write_fasta_record(out_file, header, seq)
util.remove_files('phyx.logfile')
return cleaned
def raxml_bs(fasta_file,
output_dir,
output_ext,
seq_type,
cpus,
seed,
replicates=100):
"""Build a bootstrapped tree with raxml."""
model = "PROTCATWAG" if seq_type == "aa" else "GTRCAT"
tree = util.file_name(fasta_file, output_ext)
cmd = ' '.join([
'raxml', '-T {}'.format(cpus), '-f a', '-x {}'.format(seed),
'-p {}'.format(seed), '-m {}'.format(model),
'-# {}'.format(replicates), '-s {}'.format(fasta_file),
'-n {}'.format(tree)
])
with util.cd(output_dir):
subprocess.check_call(cmd, shell=True)
tree_src = 'RAxML_bipartitions.' + tree
move(tree_src, tree)
util.remove_files('RAxML_*')
return tree