def run(self, start=None, stop=None, step=None):
"""Runs _single_universe on each system and _single_frame
on each frame in the system.
First iterates through keys of ensemble, then runs _setup_system
which defines the system and trajectory. Then iterates over
trajectory frames.
"""
logger.info("Setting up systems")
self._prepare_ensemble()
for self._key in ProgressBar(self._ensemble.keys(), verbose=True):
self._setup_system(self._key, start=start, stop=stop, step=step)
self._prepare_universe()
self._single_universe()
for i, ts in enumerate(
ProgressBar(
self._trajectory[self.start:self.stop:self.step],
verbose=True,
postfix=f'running system {self._key}')):
self._frame_index = i
self._ts = ts
self.frames[i] = ts.frame
self.times[i] = ts.time
self._single_frame()
self._conclude_universe()
logger.info("Moving to next universe")
logger.info("Finishing up")
self._conclude_ensemble()
return self
def test_verbose_disable(self, capsys):
for i in ProgressBar(list(range(10)), verbose=False):
pass
out, err = capsys.readouterr()
expected = ''
actual = err.strip().split('\r')[-1]
assert actual == expected
def test_output(self, capsys):
for i in ProgressBar(list(range(10))):
pass
out, err = capsys.readouterr()
expected = u'100%|██████████| 10/10 [00:00<00:00, 583.67it/s]'
actual = err.strip().split('\r')[-1]
assert actual[:24] == expected[:24]
def run(self, start=None, stop=None, step=None, verbose=None):
"""Perform the calculation
Parameters
----------
start : int, optional
start frame of analysis
stop : int, optional
stop frame of analysis
step : int, optional
number of frames to skip between each analysed frame
verbose : bool, optional
Turn on verbosity
"""
logger.info("Choosing frames to analyze")
# if verbose unchanged, use class default
verbose = getattr(self, '_verbose',
False) if verbose is None else verbose
self._setup_frames(self._trajectory, start, stop, step)
logger.info("Starting preparation")
self._prepare()
for i, ts in enumerate(ProgressBar(
self._trajectory[self.start:self.stop:self.step],
verbose=verbose)):
self._frame_index = i
self._ts = ts
self.frames[i] = ts.frame
self.times[i] = ts.time
# logger.info("--> Doing frame {} of {}".format(i+1, self.n_frames))
self._single_frame()
logger.info("Finishing up")
self._conclude()
return self
def _selection_serial(self, universe, selection_str):
selection = []
for ts in ProgressBar(universe.trajectory,
verbose=True,
total=universe.trajectory.n_frames):
selection.append(universe.select_atoms(selection_str))
return selection
def run(self, force=False):
"""Run all the required passes
Parameters
----------
force : bool, optional
Will overwrite previous results if they exist
"""
# if results exist, don't waste any time
if self.solution['results'] is not None and not force:
return
master_results = np.zeros_like(np.arange(self._starts[0],
self._stops[0], self._skip),
dtype=np.float32)
# for normalising later
counter = np.zeros_like(master_results, dtype=np.float32)
for i, (start,
stop) in ProgressBar(enumerate(zip(self._starts, self._stops)),
total=self.nruns,
desc="Performing run"):
# needed else trj seek thinks a np.int64 isn't an int?
results = self._single_run(int(start), int(stop))
nresults = len(results)
if nresults == len(master_results):
master_results += results
counter += 1.0
else:
master_results[:nresults] += results
counter[:nresults] += 1.0
master_results /= counter
self.solution['time'] = np.arange(
len(master_results),
dtype=np.float32) * self.u.trajectory.dt * self._skip
self.solution['results'] = master_results
def _prepare(self):
# access start index
self._u.trajectory[self.start]
# reference will be start index
self._reference = self._u.select_atoms(self._select)
self._atoms = self._u.select_atoms(self._select)
self._n_atoms = self._atoms.n_atoms
if self._mean is None:
self.mean = np.zeros(self._n_atoms*3)
self._calc_mean = True
else:
self.mean = self._mean.positions
self._calc_mean = False
if self.n_frames == 1:
raise ValueError('No covariance information can be gathered from a'
'single trajectory frame.\n')
n_dim = self._n_atoms * 3
self.cov = np.zeros((n_dim, n_dim))
self._ref_atom_positions = self._reference.positions
self._ref_cog = self._reference.center_of_geometry()
self._ref_atom_positions -= self._ref_cog
if self._calc_mean:
for ts in ProgressBar(self._u.trajectory[self.start:self.stop:self.step],
verbose=self._verbose, desc="Mean Calculation"):
if self.align:
mobile_cog = self._atoms.center_of_geometry()
mobile_atoms, old_rmsd = _fit_to(self._atoms.positions - mobile_cog,
self._ref_atom_positions,
self._atoms,
mobile_com=mobile_cog,
ref_com=self._ref_cog)
self.mean += self._atoms.positions.ravel()
self.mean /= self.n_frames
self.mean_atoms = self._atoms
self.mean_atoms.positions = self._atoms.positions
def helanal_trajectory(universe,
select="name CA",
begin=None,
finish=None,
matrix_filename="bending_matrix.dat",
origin_pdbfile="origin.pdb",
summary_filename="summary.txt",
screw_filename="screw.xvg",
tilt_filename="local_tilt.xvg",
fitted_tilt_filename="fit_tilt.xvg",
bend_filename="local_bend.xvg",
twist_filename="unit_twist.xvg",
prefix="helanal_",
ref_axis=None,
verbose=False):
"""Perform HELANAL helix analysis on all frames in `universe`.
Parameters
----------
universe : Universe
select : str (optional)
selection string that selects Calpha atoms [``"name CA"``]
begin : float (optional)
start analysing for time (ps) >= *begin*; ``None`` starts from the
beginning [``None``]
finish : float (optional)
stop analysis for time (ps) =< *finish*; ``None`` goes to the
end of the trajectory [``None``]
matrix_filename : str (optional)
Output file- bending matrix [``"bending_matrix.dat"``]
origin_pdbfile : str (optional)
Output file- origin pdb file [``"origin.pdb"``]
summary_filename : str (optional)
Output file- all of the basic data [``"summary.txt"``]
screw_filename : str (optional)
Output file- local tilts of individual residues from 2 to n-1
[``"screw.xvg"``]
tilt_filename : str (optional)
Output file- tilt of line of best fit applied to origin axes
[``"local_tilt.xvg"``]
bend_filename : str (optional)
Output file- local bend angles between successive local helix axes
[``"local_bend.xvg"``]
twist_filename : str (optional)
Output file- local unit twist between successive helix turns
[``"unit_twist.xvg"``]
prefix : str (optional)
Prefix to add to all output file names; set to ``None`` to disable
[``"helanal__"``]
ref_axis : array_like (optional)
Calculate tilt angle relative to the axis; if ``None`` then ``[0,0,1]``
is chosen [``None``]
verbose : bool (optional)
Toggle diagnostic outputs. [``True``]
Raises
------
ValueError
If the specified start (begin) time occurs after the end of the
trajectory object.
If the specified finish time precedes the specified start time or
current time stamp of trajectory object.
Notes
-----
Only a single helix is analyzed. Use the selection to specify the helix,
e.g. with "name CA and resid 1:20" or use start=1, stop=20.
.. versionchanged:: 0.13.0
New `quiet` keyword to silence frame progress output and most of the
output that used to be printed to stdout is now logged to the logger
*MDAnalysis.analysis.helanal* (at logelevel *INFO*).
.. versionchanged:: 0.16.0
Removed the `start` and `end` keywords for selecting residues because this can
be accomplished more transparently with `select`. The first and last resid
are directly obtained from the selection.
.. deprecated:: 0.16.0
The `quiet` keyword argument is deprecated in favor of the new
`verbose` one.
.. versionchanged:: 0.20.0
ProgressMeter now iterates over the number of frames analysed.
.. versionchanged:: 1.0.0
Changed `selection` keyword to `select`
"""
if ref_axis is None:
ref_axis = np.array([0., 0., 1.])
else:
# enable MDA API so that one can use a tuple of atoms or AtomGroup with
# two atoms
ref_axis = np.asarray(ref_axis)
ca = universe.select_atoms(select)
start, end = ca.resids[[0, -1]]
trajectory = universe.trajectory
# Validate user supplied begin / end times
traj_end_time = trajectory.ts.time + trajectory.totaltime
if begin is not None:
if traj_end_time < begin:
# Begin occurs after the end of the trajectory, throw error
msg = ("The input begin time ({0} ps) occurs after the end "
"of the trajectory ({1} ps)".format(begin, traj_end_time))
raise ValueError(msg)
elif trajectory.ts.time > begin:
# Begin occurs before trajectory start, warn and reset
msg = ("The input begin time ({0} ps) precedes the starting "
"trajectory time --- Setting starting frame to 0".format(
begin))
warnings.warn(msg)
logger.warning(msg)
start_frame = None
else:
start_frame = int(
np.ceil((begin - trajectory.ts.time) / trajectory.ts.dt))
else:
start_frame = None
if finish is not None:
if (begin is not None) and (begin > finish):
# finish occurs before begin time
msg = ("The input finish time ({0} ps) precedes the input begin "
"time ({1} ps)".format(finish, begin))
raise ValueError(msg)
elif trajectory.ts.time > finish:
# you'd be starting with a finish time(in ps) that has already
# passed or is not available
msg = ("The input finish time ({0} ps) precedes the current "
"trajectory time ({1} ps)".format(finish, trajectory.time))
raise ValueError(msg)
elif traj_end_time < finish:
# finish time occurs after the end of trajectory, warn
msg = ("The input finish time ({0} ps) occurs after the end of "
"the trajectory ({1} ps). Finish time will be set to the "
"end of the trajectory".format(finish, traj_end_time))
warnings.warn(msg)
logger.warning(msg)
end_frame = None
else:
# To replicate the original behaviour of break when
# trajectory.time > finish, we add 1 here.
end_frame = int(
np.floor((finish - trajectory.ts.time) // trajectory.ts.dt) +
1)
else:
end_frame = None
start_frame, end_frame, frame_step = trajectory.check_slice_indices(
start_frame, end_frame, 1)
n_frames = len(range(start_frame, end_frame, frame_step))
if start is not None and end is not None:
logger.info("Analysing from residue %d to %d", start, end)
elif start is not None and end is None:
logger.info("Analysing from residue %d to the C termini", start)
elif start is None and end is not None:
logger.info("Analysing from the N termini to %d", end)
logger.info("Analysing %d/%d residues", ca.n_atoms,
universe.atoms.n_residues)
if prefix is not None:
prefix = str(prefix)
matrix_filename = prefix + matrix_filename
origin_pdbfile = prefix + origin_pdbfile
summary_filename = prefix + summary_filename
screw_filename = prefix + screw_filename
tilt_filename = prefix + tilt_filename
fitted_tilt_filename = prefix + fitted_tilt_filename
bend_filename = prefix + bend_filename
twist_filename = prefix + twist_filename
backup_file(matrix_filename)
backup_file(origin_pdbfile)
backup_file(summary_filename)
backup_file(screw_filename)
backup_file(tilt_filename)
backup_file(fitted_tilt_filename)
backup_file(bend_filename)
backup_file(twist_filename)
global_height = []
global_twist = []
global_rnou = []
global_bending = []
global_bending_matrix = []
global_tilt = []
global_fitted_tilts = []
global_screw = []
for ts in ProgressBar(trajectory[start_frame:end_frame:frame_step],
verbose=verbose,
desc="Helix analysis"):
frame = ts.frame
ca_positions = ca.positions
twist, bending_angles, height, rnou, origins, local_helix_axes, local_screw_angles = \
main_loop(ca_positions, ref_axis=ref_axis)
origin_pdb(origins, origin_pdbfile)
#calculate local bending matrix( it is looking at all i, j combinations)
if len(global_bending_matrix) == 0:
global_bending_matrix = [[[] for item in local_helix_axes]
for item in local_helix_axes]
for i in range(len(local_helix_axes)):
for j in range(i + 1, len(local_helix_axes)):
angle = np.rad2deg(
np.arccos(np.dot(local_helix_axes[i],
local_helix_axes[j])))
global_bending_matrix[i][j].append(angle)
#global_bending_matrix[j][i].append(angle)
#global_bending_matrix[i][i].append(0.)
fit_vector, fit_tilt = vector_of_best_fit(origins)
global_height += height
global_twist += twist
global_rnou += rnou
#global_screw.append(local_screw_angles)
global_fitted_tilts.append(np.rad2deg(fit_tilt))
#print out rotations across the helix to a file
with open(twist_filename, "a") as twist_output:
print(frame, end='', file=twist_output)
for loc_twist in twist:
print(loc_twist, end='', file=twist_output)
print("", file=twist_output)
with open(bend_filename, "a") as bend_output:
print(frame, end='', file=bend_output)
for loc_bend in bending_angles:
print(loc_bend, end='', file=bend_output)
print("", file=bend_output)
with open(screw_filename, "a") as rot_output:
print(frame, end='', file=rot_output)
for rotation in local_screw_angles:
print(rotation, end='', file=rot_output)
print("", file=rot_output)
with open(tilt_filename, "a") as tilt_output:
print(frame, end='', file=tilt_output)
for tilt in local_helix_axes:
print(np.rad2deg(mdamath.angle(tilt, ref_axis)),
end='',
file=tilt_output)
print("", file=tilt_output)
with open(fitted_tilt_filename, "a") as tilt_output:
print(frame, np.rad2deg(fit_tilt), file=tilt_output)
if len(global_bending) == 0:
global_bending = [[] for item in bending_angles]
#global_tilt = [ [] for item in local_helix_axes ]
for store, tmp in zip(global_bending, bending_angles):
store.append(tmp)
#for store,tmp in zip(global_tilt,local_helix_axes): store.append(mdamath.angle(tmp,ref_axis))
twist_mean, twist_sd, twist_abdev = stats(global_twist)
height_mean, height_sd, height_abdev = stats(global_height)
rnou_mean, rnou_sd, rnou_abdev = stats(global_rnou)
ftilt_mean, ftilt_sd, ftilt_abdev = stats(global_fitted_tilts)
bending_statistics = [stats(item) for item in global_bending]
#tilt_statistics = [ stats(item) for item in global_tilt]
bending_statistics_matrix = [[stats(col) for col in row]
for row in global_bending_matrix]
with open(matrix_filename, 'w') as mat_output:
print("Mean", file=mat_output)
for row in bending_statistics_matrix:
for col in row:
formatted_angle = "{0:6.1f}".format(col[0])
print(formatted_angle, end='', file=mat_output)
print('', file=mat_output)
print('\nSD', file=mat_output)
for row in bending_statistics_matrix:
for col in row:
formatted_angle = "{0:6.1f}".format(col[1])
print(formatted_angle, end='', file=mat_output)
print('', file=mat_output)
print("\nABDEV", file=mat_output)
for row in bending_statistics_matrix:
for col in row:
formatted_angle = "{0:6.1f}".format(col[2])
print(formatted_angle, end='', file=mat_output)
print('', file=mat_output)
logger.info("Height: %g SD: %g ABDEV: %g (Angstroem)", height_mean,
height_sd, height_abdev)
logger.info("Twist: %g SD: %g ABDEV: %g", twist_mean, twist_sd,
twist_abdev)
logger.info("Residues/turn: %g SD: %g ABDEV: %g", rnou_mean, rnou_sd,
rnou_abdev)
logger.info("Fitted tilt: %g SD: %g ABDEV: %g", ftilt_mean, ftilt_sd,
ftilt_abdev)
logger.info("Local bending angles:")
residue_statistics = list(zip(*bending_statistics))
measure_names = ["Mean ", "SD ", "ABDEV"]
if start is None:
output = " ".join([
"{0:8d}".format(item)
for item in range(4,
len(residue_statistics[0]) + 4)
])
else:
output = " ".join([
"{0:8d}".format(item)
for item in range(start + 3,
len(residue_statistics[0]) + start + 3)
])
logger.info("ResID %s", output)
for measure, name in zip(residue_statistics, measure_names):
output = str(name) + " "
output += " ".join(["{0:8.1f}".format(residue) for residue in measure])
logger.info(output)
with open(summary_filename, 'w') as summary_output:
print("Height:",
height_mean,
"SD",
height_sd,
"ABDEV",
height_abdev,
'(nm)',
file=summary_output)
print("Twist:",
twist_mean,
"SD",
twist_sd,
"ABDEV",
twist_abdev,
file=summary_output)
print("Residues/turn:",
rnou_mean,
"SD",
rnou_sd,
"ABDEV",
rnou_abdev,
file=summary_output)
print("Local bending angles:", file=summary_output)
residue_statistics = list(zip(*bending_statistics))
measure_names = ["Mean ", "SD ", "ABDEV"]
print("ResID", end='', file=summary_output)
if start is None:
for item in range(4, len(residue_statistics[0]) + 4):
output = "{0:8d}".format(item)
print(output, end='', file=summary_output)
else:
for item in range(start + 3,
len(residue_statistics[0]) + start + 3):
output = "{0:8d}".format(item)
print(output, end='', file=summary_output)
print('', file=summary_output)
for measure, name in zip(residue_statistics, measure_names):
print(name, end='', file=summary_output)
for residue in measure:
output = "{0:8.1f}".format(residue)
print(output, end='', file=summary_output)
print('', file=summary_output)
