def customize(cls, input_alignment, input_partitions, output_tree, dependencies = ()):
command = AtomicCmdBuilder("raxmlHPC")
# Compute a randomized parsimony starting tree
command.set_option("-y")
# Output files are saved with a .Pypeline postfix, and subsequently renamed
command.set_option("-n", "Pypeline")
# Model required, but not used
command.set_option("-m", "GTRGAMMA")
# Ensures that output is saved to the temporary directory
command.set_option("-w", "%(TEMP_DIR)s")
# Set random seed for bootstrap generation. May be set to a fixed value to allow replicability.
command.set_option("-p", int(random.random() * 2**31 - 1), fixed = False)
# Symlink to sequence and partitions, to prevent the creation of *.reduced files outside temp folder
command.set_option("-s", "%(TEMP_OUT_ALIGNMENT)s")
command.set_option("-q", "%(TEMP_OUT_PARTITION)s")
command.set_kwargs(IN_ALIGNMENT = input_alignment,
IN_PARTITION = input_partitions,
# TEMP_OUT_ is used to automatically remove these files
TEMP_OUT_ALIGNMENT = "RAxML_alignment",
TEMP_OUT_PARTITION = "RAxML_partitions",
TEMP_OUT_INFO = "RAxML_info.Pypeline",
OUT_TREE = output_tree,
CHECK_VERSION = RAXML_VERSION)
return {"command" : command}
def customize(cls, input_alignment, input_partition, output_alignment, output_partition, dependencies = ()):
command = AtomicCmdBuilder("raxmlHPC")
# Read and (in the case of empty columns) reduce input
command.set_option("-f", "c")
# Output files are saved with a .Pypeline postfix, and subsequently renamed
command.set_option("-n", "Pypeline")
# Model required, but not used
command.set_option("-m", "GTRGAMMA")
# Ensures that output is saved to the temporary directory
command.set_option("-w", "%(TEMP_DIR)s")
# Symlink to sequence and partitions, to prevent the creation of *.reduced files outside temp folder
# In addition, it may be nessesary to remove the .reduced files if created
command.set_option("-s", "%(TEMP_IN_ALIGNMENT)s")
command.set_option("-q", "%(TEMP_IN_PARTITION)s")
command.set_kwargs(IN_ALIGNMENT = input_alignment,
IN_PARTITION = input_partition,
TEMP_IN_ALIGNMENT = "RAxML_alignment",
TEMP_IN_PARTITION = "RAxML_partitions",
TEMP_OUT_INFO = "RAxML_info.Pypeline",
OUT_ALIGNMENT = output_alignment,
OUT_PARTITION = output_partition,
CHECK_VERSION = RAXML_VERSION)
return {"command" : command}
def customize(cls, input_alignment, input_partition, template, start = 0, bootstraps = 50, dependencies = ()):
command = AtomicCmdBuilder("raxmlHPC", set_cwd = True)
# Read and (in the case of empty columns) reduce input
command.set_option("-f", "j")
# Output files are saved with a .Pypeline postfix, and subsequently renamed
command.set_option("-n", "Pypeline")
# Model required, but not used
command.set_option("-m", "GTRGAMMA")
# Set random seed for bootstrap generation. May be set to a fixed value to allow replicability.
command.set_option("-b", int(random.random() * 2**31 - 1), fixed = False)
# Generate a single bootstrap alignment (makes growing the number of bootstraps easier).
command.set_option("-N", int(bootstraps), fixed = False)
# Symlink to sequence and partitions, to prevent the creation of *.reduced files outside temp folder
# In addition, it may be nessesary to remove the .reduced files if created
command.set_option("-s", "input.alignment")
command.set_option("-q", "input.partition")
bootstrap_files = {"IN_ALIGNMENT" : input_alignment,
"IN_PARTITION" : input_partition,
"TEMP_OUT_INF" : "RAxML_info.Pypeline",
"TEMP_OUT_ALN" : "input.alignment",
"TEMP_OUT_PAR" : "input.partition",
"CHECK_VERSION": RAXML_VERSION}
for (index, (_, filename)) in enumerate(cls._bootstraps(template, bootstraps, start)):
bootstrap_files["OUT_BS_%03i" % index] = filename
command.set_kwargs(**bootstrap_files)
return {"command" : command}
def customize(cls,
input_alignment,
input_partition,
output_file,
dependencies=()):
"""
Arguments:
input_alignment -- An alignment file in a format readable by RAxML.
input_partition -- A set of partitions in a format readable by RAxML.
output_filename -- Filename for the output binary sequence."""
command = AtomicCmdBuilder("parse-examl", set_cwd=True)
command.set_option("-s", "%(TEMP_OUT_ALN)s")
command.set_option("-q", "%(TEMP_OUT_PART)s")
# Output file will be named output.binary, and placed in the CWD
command.set_option("-n", "output")
# Substitution model
command.set_option("-m", "DNA", fixed=False)
command.set_kwargs( # Auto-delete: Symlinks
TEMP_OUT_PART=os.path.basename(input_partition),
TEMP_OUT_ALN=os.path.basename(input_alignment),
# Input files, are not used directly (see below)
IN_ALIGNMENT=input_alignment,
IN_PARTITION=input_partition,
# Final output file, are not created directly
OUT_BINARY=output_file,
CHECK_EXAML=PARSER_VERSION)
return {"command": command}
def test_builder__set__kwargs__overwriting():
expected = {"IN_PATH": "/a/b/"}
builder = AtomicCmdBuilder("echo")
builder.set_kwargs(IN_PATH="/a/b/")
assert_raises(AtomicCmdBuilderError,
builder.set_kwargs, IN_PATH="/dst/file")
assert_equal(builder.kwargs, expected)
def test_builder__set_kwargs__after_finalize():
expected = {"IN_PATH": "/a/b/"}
builder = AtomicCmdBuilder("echo")
builder.set_kwargs(IN_PATH="/a/b/")
builder.finalize()
assert_raises(AtomicCmdBuilderError, builder.set_kwargs, OUT_PATH="/dst/file")
assert_equal(builder.kwargs, expected)
def customize(cls, input_alignment, input_partition, output_file, dependencies = ()):
"""
Arguments:
input_alignment -- An alignment file in a format readable by RAxML.
input_partition -- A set of partitions in a format readable by RAxML.
output_filename -- Filename for the output binary sequence."""
command = AtomicCmdBuilder("parse-examl", set_cwd = True)
command.set_option("-s", "%(TEMP_OUT_ALN)s")
command.set_option("-q", "%(TEMP_OUT_PART)s")
# Output file will be named output.binary, and placed in the CWD
command.set_option("-n", "output")
# Substitution model
command.set_option("-m", "DNA", fixed = False)
command.set_kwargs(# Auto-delete: Symlinks
TEMP_OUT_PART = os.path.basename(input_partition),
TEMP_OUT_ALN = os.path.basename(input_alignment),
# Input files, are not used directly (see below)
IN_ALIGNMENT = input_alignment,
IN_PARTITION = input_partition,
# Final output file, are not created directly
OUT_BINARY = output_file,
CHECK_EXAML = PARSER_VERSION)
return {"command" : command}
def _bowtie2_template(call, prefix, iotype = "IN", **kwargs):
params = AtomicCmdBuilder(call, **kwargs)
for postfix in ("1.bt2", "2.bt2", "3.bt2", "4.bt2", "rev.1.bt2", "rev.2.bt2"):
key = "%s_PREFIX_%s" % (iotype, postfix.upper())
params.set_kwargs(**{key : (prefix + "." + postfix)})
return params
def test_builder__set__kwargs__overwriting():
expected = {"IN_PATH": "/a/b/"}
builder = AtomicCmdBuilder("echo")
builder.set_kwargs(IN_PATH="/a/b/")
assert_raises(AtomicCmdBuilderError,
builder.set_kwargs,
IN_PATH="/dst/file")
assert_equal(builder.kwargs, expected)
def customize(cls, input_file, output_file, algorithm = "auto", dependencies = ()):
command = AtomicCmdBuilder(_PRESETS[algorithm.lower()])
command.add_value("%(IN_FASTA)s")
command.set_kwargs(IN_FASTA = input_file,
OUT_STDOUT = output_file,
CHECK_VERSION = MAFFT_VERSION)
return {"command" : command,
"dependencies" : dependencies}
def test_builder__set_kwargs__after_finalize():
expected = {"IN_PATH": "/a/b/"}
builder = AtomicCmdBuilder("echo")
builder.set_kwargs(IN_PATH="/a/b/")
builder.finalize()
assert_raises(AtomicCmdBuilderError,
builder.set_kwargs,
OUT_PATH="/dst/file")
assert_equal(builder.kwargs, expected)
def customize(cls, input_file, output_file, algorithm = "auto", dependencies = ()):
command = AtomicCmdBuilder(_PRESETS[algorithm.lower()])
command.add_value("%(IN_FASTA)s")
command.set_kwargs(IN_FASTA = input_file,
OUT_STDOUT = output_file,
CHECK_VERSION = MAFFT_VERSION)
return {"command" : command,
"dependencies" : dependencies}
def _get_bwa_template(call, prefix, iotype="IN", **kwargs):
extensions = ["amb", "ann", "bwt", "pac", "sa"]
try:
if BWA_VERSION.version < (0, 6, 0):
extensions.extend(("rbwt", "rpac", "rsa"))
except versions.VersionRequirementError:
pass # Ignored here, handled elsewhere
params = AtomicCmdBuilder(call, **kwargs)
for postfix in extensions:
key = "%s_PREFIX_%s" % (iotype, postfix.upper())
params.set_kwargs(**{key: (prefix + "." + postfix)})
return params
def _get_bwa_template(call, prefix, iotype="IN", **kwargs):
extensions = ["amb", "ann", "bwt", "pac", "sa"]
try:
if BWA_VERSION.version < (0, 6, 0):
extensions.extend(("rbwt", "rpac", "rsa"))
except versions.VersionRequirementError:
pass # Ignored here, handled elsewhere
params = AtomicCmdBuilder(call, **kwargs)
for postfix in extensions:
key = "%s_PREFIX_%s" % (iotype, postfix.upper())
params.set_kwargs(**{key: (prefix + "." + postfix)})
return params
def __init__(self,
input_alignment,
input_partition,
output_file,
dependencies=()):
"""
Arguments:
input_alignment -- An alignment file in a format readable by RAxML.
input_partition -- A set of partitions in a format readable by RAxML.
output_filename -- Filename for the output binary sequence."""
command = AtomicCmdBuilder("parse-examl", set_cwd=True)
command.set_option("-s", "%(TEMP_OUT_ALN)s")
command.set_option("-q", "%(TEMP_OUT_PART)s")
# Output file will be named output.binary, and placed in the CWD
command.set_option("-n", "output")
# Substitution model
command.set_option("-m", "DNA", fixed=False)
command.set_kwargs( # Auto-delete: Symlinks
TEMP_OUT_PART=os.path.basename(input_partition),
TEMP_OUT_ALN=os.path.basename(input_alignment),
# Input files, are not used directly (see below)
IN_ALIGNMENT=input_alignment,
IN_PARTITION=input_partition,
# Final output file, are not created directly
OUT_BINARY=output_file,
CHECK_EXAML=PARSER_VERSION,
)
CommandNode.__init__(
self,
command=command.finalize(),
description="<ExaMLParser: '%s' -> '%s'>" %
(input_alignment, output_file),
dependencies=dependencies,
)
self._symlinks = [
os.path.abspath(input_alignment),
os.path.abspath(input_partition),
]
self._output_file = os.path.basename(output_file)
def test_builder__finalize__calls_atomiccmd():
was_called = []
class _AtomicCmdMock:
def __init__(self, *args, **kwargs):
assert_equal(args, (["echo", "-out", "%(OUT_FILE)s", "%(IN_FILE)s"],))
assert_equal(kwargs, {"IN_FILE": "/in/file",
"OUT_FILE": "/out/file",
"set_cwd": True})
was_called.append(True)
with Monkeypatch("paleomix.atomiccmd.builder.AtomicCmd", _AtomicCmdMock):
builder = AtomicCmdBuilder("echo", set_cwd=True)
builder.add_option("-out", "%(OUT_FILE)s")
builder.add_value("%(IN_FILE)s")
builder.set_kwargs(OUT_FILE="/out/file",
IN_FILE="/in/file")
builder.finalize()
assert was_called
def test_builder__finalize__calls_atomiccmd():
was_called = []
class _AtomicCmdMock(object):
def __init__(self, *args, **kwargs):
assert_equal(args,
(["echo", "-out", "%(OUT_FILE)s", "%(IN_FILE)s"], ))
assert_equal(kwargs, {
"IN_FILE": "/in/file",
"OUT_FILE": "/out/file",
"set_cwd": True
})
was_called.append(True)
with Monkeypatch("paleomix.atomiccmd.builder.AtomicCmd", _AtomicCmdMock):
builder = AtomicCmdBuilder("echo", set_cwd=True)
builder.add_option("-out", "%(OUT_FILE)s")
builder.add_value("%(IN_FILE)s")
builder.set_kwargs(OUT_FILE="/out/file", IN_FILE="/in/file")
builder.finalize()
assert was_called
def customize(cls, input_alignment, output_tree, dependencies=()):
"""
Arguments:
input_alignment -- An alignment file in a format readable by RAxML.
output_tree -- Filename for the output newick tree."""
command = AtomicCmdBuilder("parsimonator", set_cwd=True)
command.set_option("-s", "%(TEMP_OUT_ALN)s")
command.set_option("-n", "output")
# Random seed for the stepwise addition process
command.set_option("-p", int(random.random() * 2**31 - 1), fixed=False)
command.set_kwargs( # Auto-delete: Symlinks
TEMP_OUT_ALN=os.path.basename(input_alignment),
# Input files, are not used directly (see below)
IN_ALIGNMENT=input_alignment,
# Final output file, are not created directly
OUT_TREE=output_tree)
return {"command": command}
def customize(cls, input_alignment, output_tree, dependencies = ()):
"""
Arguments:
input_alignment -- An alignment file in a format readable by RAxML.
output_tree -- Filename for the output newick tree."""
command = AtomicCmdBuilder("parsimonator", set_cwd = True)
command.set_option("-s", "%(TEMP_OUT_ALN)s")
command.set_option("-n", "output")
# Random seed for the stepwise addition process
command.set_option("-p", int(random.random() * 2**31 - 1), fixed = False)
command.set_kwargs(# Auto-delete: Symlinks
TEMP_OUT_ALN = os.path.basename(input_alignment),
# Input files, are not used directly (see below)
IN_ALIGNMENT = input_alignment,
# Final output file, are not created directly
OUT_TREE = output_tree)
return {"command" : command}
def test_builder__set_kwargs__called_twice():
expected = {"IN_PATH": "/a/b/", "OUT_PATH": "/dst/file"}
builder = AtomicCmdBuilder("echo")
builder.set_kwargs(OUT_PATH="/dst/file")
builder.set_kwargs(IN_PATH="/a/b/")
assert_equal(builder.kwargs, expected)
def test_builder__set_kwargs__called_twice():
expected = {"IN_PATH": "/a/b/", "OUT_PATH": "/dst/file"}
builder = AtomicCmdBuilder("echo")
builder.set_kwargs(OUT_PATH="/dst/file")
builder.set_kwargs(IN_PATH="/a/b/")
assert_equal(builder.kwargs, expected)
def customize(cls, input_alignment, output_template, input_partition=None,
threads=1, dependencies=()):
"""
Arguments:
input_alignment -- An alignment file in a format readable by RAxML.
input_partition -- A set of partitions in a format readable by RAxML.
output_template -- A template string used to construct final filenames. Should consist
of a full path, including a single '%s', which is replaced with the
variable part of RAxML output files (e.g. 'info', 'bestTree', ...).
Example destination: '/disk/project/SN013420.RAxML.%s'
Example output: '/disk/project/SN013420.RAxML.bestTree'
"""
if threads > 1:
command = AtomicCmdBuilder("raxmlHPC-PTHREADS")
command.set_option("-T", threads)
version = RAXML_PTHREADS_VERSION
else:
command = AtomicCmdBuilder("raxmlHPC")
version = RAXML_VERSION
# Perform rapid bootstrapping
command.set_option("-f", "a")
# Output files are saved with a .PALEOMIX postfix, and subsequently renamed
command.set_option("-n", "PALEOMIX")
# Ensures that output is saved to the temporary directory
command.set_option("-w", "%(TEMP_DIR)s")
# Symlink to sequence and partitions, to prevent the creation of *.reduced files outside temp folder
# In addition, it may be nessesary to remove the .reduced files if created
command.set_option("-s", "%(TEMP_OUT_ALN)s")
if input_partition is not None:
command.set_option("-q", "%(TEMP_OUT_PART)s")
command.set_kwargs(IN_PARTITION=input_partition,
TEMP_OUT_PART=os.path.basename(input_partition),
TEMP_OUT_PART_R=os.path.basename(input_partition) + ".reduced")
command.set_kwargs( # Auto-delete: Symlinks and .reduced files that RAxML may generate
TEMP_OUT_ALN=os.path.basename(input_alignment),
TEMP_OUT_ALN_R=os.path.basename(input_alignment) + ".reduced",
# Input files, are not used directly (see below)
IN_ALIGNMENT=input_alignment,
# Final output files, are not created directly
OUT_INFO=output_template % "info",
OUT_BESTTREE=output_template % "bestTree",
OUT_BOOTSTRAP=output_template % "bootstrap",
OUT_BIPART=output_template % "bipartitions",
OUT_BIPARTLABEL=output_template % "bipartitionsBranchLabels",
CHECK_VERSION=version)
# Use the GTRGAMMA model of NT substitution by default
command.set_option("-m", "GTRGAMMAI", fixed=False)
# Enable Rapid Boostrapping and set random seed. May be set to a fixed value to allow replicability.
command.set_option("-x", int(random.random() * 2**31 - 1), fixed=False)
# Set random seed for parsimony inference. May be set to a fixed value to allow replicability.
command.set_option("-p", int(random.random() * 2**31 - 1), fixed=False)
# Terminate bootstrapping upon convergence, rather than after a fixed number of repetitions
command.set_option("-N", "autoMRE", fixed=False)
return {"command": command}
def __init__(
self,
input_alignment,
output_template,
input_partition=None,
model="GTRGAMMAI",
replicates="autoMRE",
threads=1,
dependencies=(),
):
"""
Arguments:
input_alignment -- An alignment file in a format readable by RAxML.
input_partition -- A set of partitions in a format readable by RAxML.
output_template -- A template string used to construct final filenames. Should
consist of a full path, including a single '%s', which is
replaced with the variable part of RAxML output files (e.g.
'info', 'bestTree', ...).
Example destination: '/disk/project/SN013420.RAxML.%s'
Example output: '/disk/project/SN013420.RAxML.bestTree'
"""
if threads > 1:
command = AtomicCmdBuilder("raxmlHPC-PTHREADS")
command.set_option("-T", threads)
version = RAXML_PTHREADS_VERSION
else:
command = AtomicCmdBuilder("raxmlHPC")
version = RAXML_VERSION
# Perform rapid bootstrapping
command.set_option("-f", "a")
# Output files are saved with a .PALEOMIX postfix, and subsequently renamed
command.set_option("-n", "PALEOMIX")
# Ensures that output is saved to the temporary directory
command.set_option("-w", "%(TEMP_DIR)s")
# Symlink to sequence and partitions, to prevent the creation of *.reduced files
# outside temp folder. In addition, it may be nessesary to remove the .reduced
# files if created
command.set_option("-s", "%(TEMP_OUT_ALN)s")
if input_partition is not None:
command.set_option("-q", "%(TEMP_OUT_PART)s")
command.set_kwargs(
IN_PARTITION=input_partition,
TEMP_OUT_PART=os.path.basename(input_partition),
TEMP_OUT_PART_R=os.path.basename(input_partition) + ".reduced",
)
command.set_kwargs(
# Auto-delete: Symlinks and .reduced files that RAxML may generate
TEMP_OUT_ALN=os.path.basename(input_alignment),
TEMP_OUT_ALN_R=os.path.basename(input_alignment) + ".reduced",
# Input files, are not used directly (see below)
IN_ALIGNMENT=input_alignment,
# Final output files, are not created directly
OUT_INFO=output_template % "info",
OUT_BESTTREE=output_template % "bestTree",
OUT_BOOTSTRAP=output_template % "bootstrap",
OUT_BIPART=output_template % "bipartitions",
OUT_BIPARTLABEL=output_template % "bipartitionsBranchLabels",
CHECK_VERSION=version,
)
# Use the GTRGAMMA model of NT substitution by default
command.set_option("-m", model, fixed=False)
# Enable Rapid Boostrapping and set random seed. May be set to a fixed value to
# allow replicability.
command.set_option("-x", int(random.random() * 2 ** 31 - 1), fixed=False)
# Set random seed for parsimony inference. May be set to allow replicability.
command.set_option("-p", int(random.random() * 2 ** 31 - 1), fixed=False)
# Terminate bootstrapping upon convergence, not after N repetitions
command.set_option("-N", replicates, fixed=False)
self._symlinks = [input_alignment, input_partition]
self._template = os.path.basename(output_template)
CommandNode.__init__(
self,
command=command.finalize(),
description="<RAxMLRapidBS: '%s' -> '%s'>"
% (input_alignment, output_template % ("*",)),
threads=threads,
dependencies=dependencies,
)