def run(self, input_tree, msa_file, num_replicates, model, base_type, frac,
output_dir):
"""Bootstrap multiple sequence alignment.
Parameters
----------
input_tree : str
Tree inferred with all data.
msa_file : str
File containing multiple sequence alignment for all taxa.
num_replicates : int
Number of replicates to perform.
model : str
Desired model of evolution.
base_type : str
Indicates if bases are nucleotides or amino acids.
frac : float
Fraction of alignment to subsample.
output_dir : str
Directory for bootstrap trees.
"""
assert (model in ['wag', 'lg', 'jtt'])
assert (base_type in ['nt', 'prot'])
self.model = model
self.base_type = base_type
self.frac = frac
self.replicate_dir = os.path.join(output_dir, 'replicates')
make_sure_path_exists(self.replicate_dir)
# read full multiple sequence alignment
self.msa = seq_io.read(msa_file)
# calculate replicates
self.logger.info('Calculating bootstrap replicates:')
parallel = Parallel(self.cpus)
parallel.run(self._producer, None, range(num_replicates),
self._progress)
# calculate support values
rep_tree_files = []
for rep_index in range(num_replicates):
rep_tree_files.append(
os.path.join(self.replicate_dir,
'bootstrap_tree.r_' + str(rep_index) + '.tree'))
output_tree = os.path.join(
output_dir,
remove_extension(input_tree) + '.bootstrap.tree')
bootstrap_support(input_tree, rep_tree_files, output_tree)
return output_tree
def bootstrap(self, input_tree, msa_file, seq_type, model_str, gamma,
num_replicates, output_dir, cpus):
"""Perform non-parametric bootstrapping.
Parameters
----------
input_tree : str
File containing newick tree to decorate with bootstraps.
msa_file : str
Fasta file containing multiple sequence alignment.
seq_type : str
Specifies multiple sequences alignment is of 'nt' or 'prot'.
model_str : str
Specified either the 'wag' or 'jtt' model.
gamma : bool
Indicates if GAMMA model should be used
num_replicates : int
Number of replicates to perform.
output_dir: str
Output directory to contain bootstrap trees.
cpus : int
Number of cpus to use.
"""
assert (seq_type.upper() in ['NT', 'PROT'])
assert (model_str.upper() in ['WAG', 'LG', 'JTT', 'GTR'])
self.output_dir = output_dir
self.seq_type = seq_type
self.model = model_str
self.gamma = gamma
self.msa = seq_io.read(msa_file)
# calculate replicates
parallel = Parallel(cpus)
replicate_numbers = list(range(num_replicates))
parallel.run(self._bootstrap, None, replicate_numbers, None)
# calculate support values
rep_tree_files = []
for rep_index in replicate_numbers:
rep_tree_files.append(
os.path.join(self.output_dir, 'rep_%d' % rep_index,
'bootstrap.tree'))
tree_name = os.path.splitext(os.path.basename(input_tree))[0]
output_tree = os.path.join(output_dir, tree_name + '.bootstrap.tree')
bootstrap_support(input_tree, rep_tree_files, output_tree)
return output_tree
def bootstrap(self, input_tree, msa_file, model_str, num_replicates, output_dir, cpus):
"""Perform non-parametric bootstrapping.
Parameters
----------
input_tree : str
File containing newick tree to decorate with bootstraps.
msa_file : str
Fasta file containing multiple sequence alignment.
model_str : str
Specified either the 'WAG' or 'LG' model.
num_replicates : int
Number of replicates to perform.
output_dir: str
Output directory to contain bootstrap trees.
cpus : int
Number of cpus to use.
"""
check_on_path('seqmagick')
assert(model_str.upper() in ['WAG', 'LG'])
self.output_dir = output_dir
self.model = model_str
self.msa = seq_io.read(msa_file)
# calculate replicates
parallel = Parallel(cpus)
replicate_numbers = list(range(num_replicates))
parallel.run(self._bootstrap, None, replicate_numbers, None)
# calculate support values
rep_tree_files = []
for rep_index in replicate_numbers:
rep_tree_files.append(os.path.join(output_dir, 'rep_%d' % rep_index, 'RAxML_bestTree.support'))
tree_name = os.path.splitext(os.path.basename(input_tree))[0]
output_tree = os.path.join(output_dir, tree_name + '.bootstrap.tree')
bootstrap_support(input_tree, rep_tree_files, output_tree)
return output_tree
def bootstrap(self, input_tree, msa_file, seq_type, model_str, num_replicates, output_tree, cpus):
"""Perform non-parametric bootstrapping.
Parameters
----------
input_tree : str
File containing newick tree to decorate with bootstraps.
msa_file : str
Fasta file containing multiple sequence alignment.
seq_type : str
Specifies multiple sequences alignment is of 'nt' or 'prot'.
model_str : str
Specified either the 'wag' or 'jtt' model.
num_replicates : int
Number of replicates to perform.
output_tree: str
Output file containing tree with bootstrap values.
cpus : int
Number of cpus to use.
"""
assert(seq_type in ['nt', 'prot'])
assert(model_str in ['wag', 'jtt'])
self.replicate_dir = tempfile.mkdtemp()
self.seq_type = seq_type
self.model = model_str
self.msa = seq_io.read(msa_file)
# calculate replicates
parallel = Parallel(cpus)
parallel.run(self._bootstrap, None, xrange(num_replicates), None)
# calculate support values
rep_tree_files = []
for rep_index in xrange(num_replicates):
rep_tree_files.append(os.path.join(self.replicate_dir, 'bootstrap.tree.' + str(rep_index) + '.tre'))
bootstrap_support(input_tree, rep_tree_files, output_tree)
shutil.rmtree(self.replicate_dir)
def run(self,
input_tree,
msa_file,
num_replicates,
model,
gamma,
base_type,
frac,
boot_dir,
output_dir):
"""Bootstrap multiple sequence alignment.
Parameters
----------
input_tree : str
Tree inferred with all data.
msa_file : str
File containing multiple sequence alignment for all taxa.
num_replicates : int
Number of replicates to perform.
model : str
Desired model of evolution.
base_type : str
Indicates if bases are nucleotides or amino acids.
frac : float
Fraction of alignment to subsample.
output_dir : str
Directory for bootstrap trees.
"""
assert(model in ['wag', 'lg', 'jtt'])
assert(base_type in ['nt', 'prot'])
self.model = model
self.gamma = gamma
self.base_type = base_type
self.frac = frac
rep_tree_files = []
if not boot_dir:
self.replicate_dir = os.path.join(output_dir, 'replicates')
make_sure_path_exists(self.replicate_dir)
# read full multiple sequence alignment
self.msa = seq_io.read(msa_file)
# calculate replicates
self.logger.info('Calculating bootstrap replicates:')
parallel = Parallel(self.cpus)
parallel.run(self._producer, None, xrange(num_replicates), self._progress)
for rep_index in xrange(num_replicates):
rep_tree_files.append(os.path.join(self.replicate_dir, 'bootstrap_tree.r_' + str(rep_index) + '.tree'))
else:
for f in os.listdir(boot_dir):
if f.endswith('.tree') or f.endswith('.tre'):
rep_tree_files.append(os.path.join(boot_dir, f))
self.logger.info('Read %d bootstrap replicates.' % len(rep_tree_files))
# calculate support values
self.logger.info('Calculating bootstrap support values.')
output_tree = os.path.join(output_dir, remove_extension(input_tree) + '.bootstrap.tree')
bootstrap_support(input_tree, rep_tree_files, output_tree)
return output_tree
def run(self, input_tree, msa_file, marker_info_file, mask_file,
perc_markers_to_keep, num_replicates, model, jk_dir, output_dir):
"""Jackknife marker genes.
Marker file should have the format:
<marker id>\t<marker name>\t<marker desc>\t<length>\n
Parameters
----------
input_tree : str
Tree inferred with all data.
msa_file : str
File containing multiple sequence alignment for all taxa.
marker_info_file : str
File indicating database id, HMM name, description and length of each marker in the alignment.
mask_file : str
File indicating masking of multiple sequence alignment.
perc_markers_to_keep : float [0, 1]
Percentage of marker genes to keep in each replicate.
num_replicates : int
Number of replicates to perform.
model : str
Desired model of evolution.
output_dir : str
Output directory for jackkife trees.
"""
assert (model in ['wag', 'jtt'])
self.model = model
self.perc_markers_to_keep = perc_markers_to_keep
# determine length of each marker gene in alignment
rep_tree_files = []
if not jk_dir:
self.replicate_dir = os.path.join(output_dir, 'replicates')
make_sure_path_exists(self.replicate_dir)
marker_lengths = []
total_len = 0
with open(marker_info_file) as f:
f.readline()
for line in f:
line_split = line.split('\t')
ml = int(line_split[3])
marker_lengths.append(ml)
total_len += ml
self.logger.info('Concatenated length of markers: %d' % total_len)
# read mask
mask = open(mask_file).readline().strip()
start = 0
self.marker_lengths = []
total_mask_len = 0
for ml in marker_lengths:
end = start + ml
zeros = mask[start:end].count('0')
start = end
self.marker_lengths.append(ml - zeros)
total_mask_len += ml - zeros
self.logger.info('Concatenated length of filtered MSA: %d' %
total_mask_len)
# read full multiple sequence alignment
self.msa = seq_io.read(msa_file)
if len(self.msa.values()[0]) != total_mask_len:
self.logger.error(
'Length of MSA does not meet length of mask.')
sys.exit()
# calculate replicates
self.logger.info('Calculating jackknife marker replicates:')
parallel = Parallel(self.cpus)
parallel.run(self._producer, None, xrange(num_replicates),
self._progress)
# calculate support
self.logger.info('Calculating support for %d replicates.' %
num_replicates)
for rep_index in xrange(num_replicates):
rep_tree_files.append(
os.path.join(self.replicate_dir,
'jk_markers.tree.' + str(rep_index) + '.tre'))
else:
for f in os.listdir(jk_dir):
if f.endswith('.tree') or f.endswith('.tre'):
rep_tree_files.append(os.path.join(jk_dir, f))
self.logger.info('Read %d jackknife replicates.' %
len(rep_tree_files))
output_tree = os.path.join(
output_dir,
remove_extension(input_tree) + '.jk_markers.tree')
bootstrap_support(input_tree, rep_tree_files, output_tree)
return output_tree
def run(self, input_tree,
msa_file,
marker_info_file,
mask_file,
perc_markers_to_keep,
num_replicates,
model,
jk_dir,
output_dir):
"""Jackknife marker genes.
Marker file should have the format:
<marker id>\t<marker name>\t<marker desc>\t<length>\n
Parameters
----------
input_tree : str
Tree inferred with all data.
msa_file : str
File containing multiple sequence alignment for all taxa.
marker_info_file : str
File indicating database id, HMM name, description and length of each marker in the alignment.
mask_file : str
File indicating masking of multiple sequence alignment.
perc_markers_to_keep : float [0, 1]
Percentage of marker genes to keep in each replicate.
num_replicates : int
Number of replicates to perform.
model : str
Desired model of evolution.
output_dir : str
Output directory for jackkife trees.
"""
assert(model in ['wag', 'jtt'])
self.model = model
self.perc_markers_to_keep = perc_markers_to_keep
# determine length of each marker gene in alignment
rep_tree_files = []
if not jk_dir:
self.replicate_dir = os.path.join(output_dir, 'replicates')
make_sure_path_exists(self.replicate_dir)
marker_lengths = []
total_len = 0
with open(marker_info_file) as f:
f.readline()
for line in f:
line_split = line.split('\t')
ml = int(line_split[3])
marker_lengths.append(ml)
total_len += ml
self.logger.info('Concatenated length of markers: %d' % total_len)
# read mask
mask = open(mask_file).readline().strip()
start = 0
self.marker_lengths = []
total_mask_len = 0
for ml in marker_lengths:
end = start + ml
zeros = mask[start:end].count('0')
start = end
self.marker_lengths.append(ml - zeros)
total_mask_len += ml - zeros
self.logger.info('Concatenated length of filtered MSA: %d' % total_mask_len)
# read full multiple sequence alignment
self.msa = seq_io.read(msa_file)
if len(self.msa.values()[0]) != total_mask_len:
self.logger.error('Length of MSA does not meet length of mask.')
sys.exit()
# calculate replicates
self.logger.info('Calculating jackknife marker replicates:')
parallel = Parallel(self.cpus)
parallel.run(self._producer, None, xrange(num_replicates), self._progress)
# calculate support
self.logger.info('Calculating support for %d replicates.' % num_replicates)
for rep_index in xrange(num_replicates):
rep_tree_files.append(os.path.join(self.replicate_dir, 'jk_markers.tree.' + str(rep_index) + '.tre'))
else:
for f in os.listdir(jk_dir):
if f.endswith('.tree') or f.endswith('.tre'):
rep_tree_files.append(os.path.join(jk_dir, f))
self.logger.info('Read %d jackknife replicates.' % len(rep_tree_files))
output_tree = os.path.join(output_dir, remove_extension(input_tree) + '.jk_markers.tree')
bootstrap_support(input_tree, rep_tree_files, output_tree)
return output_tree