def test_distances_t_at_0(get_fn):
times = np.array([[0, 0]], dtype=np.int32)
a = compute_distances_t(ptraj, pairs, times, periodic=True, opt=True)
b = compute_distances_t(ptraj, pairs, times, periodic=True, opt=False)
c = compute_distances(ptraj[:1], pairs, periodic=True, opt=True)
d = compute_distances(ptraj[:1], pairs, periodic=True, opt=False)
eq(a, c)
eq(b, d)
def __init__(self, directory="/home/harrigan/projects/msmbuilder/Tutorial/"):
# Load from disk
print "Loading trajectories"
traj_list_in = get_data.get_from_fs(os.path.join(directory, 'XTC'), os.path.join(directory, 'native.pdb'))
# Compute dihedrals
traj_list_angles = list()
traj_list = list()
print "Computing Dihedrals"
for traj in traj_list_in:
# Compute actual angles
diheds = dihedral.compute_dihedrals(traj, Dipeptide.INDICES, opt=False)
# Compute sin and cos
num_diheds = diheds.shape[1]
per_diheds = np.zeros((diheds.shape[0], num_diheds * 2))
per_diheds[:, 0:num_diheds] = np.sin(diheds)
per_diheds[:, num_diheds:num_diheds * 2] = np.cos(diheds)
distances = distance.compute_distances(traj, Dipeptide.PAIRS, periodic=True, opt=False)
distances = np.log(distances)
# Save
traj_list.append(distances)
traj_list_angles.append(diheds)
# traj_list.append(per_diheds)
self.traj_list = traj_list
self.traj_list_angles = traj_list_angles
def test_2p():
a = compute_distances(ptraj, pairs, periodic=True, opt=False)
b = compute_displacements(ptraj, pairs, periodic=True, opt=False)
assert a.shape == (len(ptraj), len(pairs))
assert b.shape == (len(ptraj), len(pairs), 3), str(b.shape)
b = np.sqrt(np.sum(np.square(b), axis=2))
eq(a, b, decimal=5)
def test_2p():
a = compute_distances(ptraj, pairs, periodic=True, opt=False)
b = compute_displacements(ptraj, pairs, periodic=True, opt=False)
assert a.shape == (len(ptraj), len(pairs))
assert b.shape == (len(ptraj), len(pairs), 3), str(b.shape)
b = np.sqrt(np.sum(np.square(b), axis=2))
eq(a, b, decimal=5)
def get_redundant_internal_coordinates(trajectory, **kwargs):
"""Compute internal coordinates from the cartesian coordinates
This extracts all of the bond lengths, bond angles and dihedral angles
from every frame in a trajectory.
Parameters
----------
trajectory : mdtraj.Trajectory
Trajectory object containing the internal coordinates
Other Parameters
----------------
ibonds : np.ndarray, optional, shape[n_bonds, 2], dtype=int
Each row gives the indices of two atoms involved in a bond
iangles : np.ndarray, optional shape[n_angles, 3], dtype=int
Each row gives the indices of three atoms which together make an angle
idihedrals : np.ndarray, optional, shape[n_dihedrals, 4], dtype=int
Each row gives the indices of the four atoms which together make a
dihedral
Notes
-----
ibonds, iangles, and idihedrals will be computed usig the first
frame in the trajectory, if not supplied
Returns
-------
internal_coords : np.ndarray, shape=[n_frames, n_bonds+n_angles+n_dihedrals]
All of the internal coordinates collected into a big array, such that
internal_coords[i,j] gives the jth coordinate for the ith frame.
"""
if 'ibonds' in kwargs and 'iangles' in kwargs and 'idihedrals' in kwargs:
ibonds = kwargs['ibonds']
iangles = kwargs['iangles']
idihedrals = kwargs['idihedrals']
else:
ibonds, iangles, idihedrals = get_connectivity(trajectory)
# convert everything to the right shape and C ordering, since
# all of these methods are in C and are going to need things to be
# the right type. The methods will all do a copy for things that
# aren't the right type, but hopefully we can only do the copy once
# instead of three times if xyzlist really does need to be reordered
# in memory
xyzlist = np.array(trajectory.xyz, dtype=np.float32, order='c')
ibonds = np.array(ibonds, dtype=np.int32, order='c')
iangles = np.array(iangles, dtype=np.int32, order='c')
idihedrals = np.array(idihedrals, dtype=np.int32, order='c')
b = compute_distances(xyzlist, ibonds)
a = compute_angles(xyzlist, iangles)
d = compute_dihedrals(xyzlist, idihedrals, degrees=False)
return np.hstack((b, a, d))
def get_redundant_internal_coordinates(trajectory, **kwargs):
"""Compute internal coordinates from the cartesian coordinates
This extracts all of the bond lengths, bond angles and dihedral angles
from every frame in a trajectory.
Parameters
----------
trajectory : mdtraj.Trajectory
Trajectory object containing the internal coordinates
Other Parameters
----------------
ibonds : np.ndarray, optional, shape[n_bonds, 2], dtype=int
Each row gives the indices of two atoms involved in a bond
iangles : np.ndarray, optional shape[n_angles, 3], dtype=int
Each row gives the indices of three atoms which together make an angle
idihedrals : np.ndarray, optional, shape[n_dihedrals, 4], dtype=int
Each row gives the indices of the four atoms which together make a
dihedral
Notes
-----
ibonds, iangles, and idihedrals will be computed usig the first
frame in the trajectory, if not supplied
Returns
-------
internal_coords : np.ndarray, shape=[n_frames, n_bonds+n_angles+n_dihedrals]
All of the internal coordinates collected into a big array, such that
internal_coords[i,j] gives the jth coordinate for the ith frame.
"""
if 'ibonds' in kwargs and 'iangles' in kwargs and 'idihedrals' in kwargs:
ibonds = kwargs['ibonds']
iangles = kwargs['iangles']
idihedrals = kwargs['idihedrals']
else:
ibonds, iangles, idihedrals = get_connectivity(trajectory)
# convert everything to the right shape and C ordering, since
# all of these methods are in C and are going to need things to be
# the right type. The methods will all do a copy for things that
# aren't the right type, but hopefully we can only do the copy once
# instead of three times if xyzlist really does need to be reordered
# in memory
xyzlist = np.array(trajectory.xyz, dtype=np.float32, order='c')
ibonds = np.array(ibonds, dtype=np.int32, order='c')
iangles = np.array(iangles, dtype=np.int32, order='c')
idihedrals = np.array(idihedrals, dtype=np.int32, order='c')
b = compute_distances(xyzlist, ibonds)
a = compute_angles(xyzlist, iangles)
d = compute_dihedrals(xyzlist, idihedrals, degrees=False)
return np.hstack((b, a, d))
def compute_rdf(traj, pairs=None, r_range=None, bin_width=0.005,
periodic=True, opt=True):
"""Compute radial distribution functions for pairs in every frame.
Parameters
----------
traj : Trajectory
Trajectory to compute radial distribution function in.
pairs : array-like, shape=(n_pairs, 2), dtype=int, optional, default=None
Each row gives the indices of two atoms.
r_range : array-like, shape=(2,), optional, default=(0.0, 1.0)
Minimum and maximum radii.
bin_width : int, optional, default=0.005
Width of the bins in nanometers.
periodic : bool, default=True
If `periodic` is True and the trajectory contains unitcell
information, we will compute distances under the minimum image
convention.
opt : bool, default=True
Use an optimized native library to compute the pair wise distances.
Returns
-------
r : np.ndarray, shape=(np.diff(r_range) / bin_width - 1), dtype=float
Radii values corresponding to the centers of the bins.
g_r : np.ndarray, shape=(np.diff(r_range) / bin_width - 1), dtype=float
Radial distribution function values at r.
See also
--------
Topology.select_pairs
"""
if r_range is None:
r_range = np.array([0.0, 1.0])
r_range = ensure_type(r_range, dtype=np.float64, ndim=1, name='r_range',
shape=(2,), warn_on_cast=False)
bins = np.arange(r_range[0], r_range[1], bin_width)
distances = compute_distances(traj, pairs, periodic=periodic, opt=opt)
g_r, edges = np.histogram(distances, bins=bins)
r = 0.5 * (edges[1:] + edges[:-1])
# Normalize by volume of the spherical shell.
# See discussion https://github.com/mdtraj/mdtraj/pull/724. There might be
# a less biased way to accomplish this. The conclusion was that this could
# be interesting to try, but is likely not hugely consequential. This method
# of doing the calculations matches the implementation in other packages like
# AmberTools' cpptraj and gromacs g_rdf.
V = (4 / 3) * np.pi * (np.power(edges[1:], 3) - np.power(edges[:-1], 3))
norm = len(pairs) * np.sum(1.0 / traj.unitcell_volumes) * V
g_r = g_r.astype(np.float64) / norm # From int64.
return r, g_r
def __init__(self,
directory="/home/harrigan/projects/msmbuilder/Tutorial/"):
# Load from disk
print "Loading trajectories"
traj_list_in = get_data.get_from_fs(
os.path.join(directory, 'XTC'),
os.path.join(directory, 'native.pdb'))
# Compute dihedrals
traj_list_angles = list()
traj_list = list()
print "Computing Dihedrals"
for traj in traj_list_in:
# Compute actual angles
diheds = dihedral.compute_dihedrals(traj,
Dipeptide.INDICES,
opt=False)
# Compute sin and cos
num_diheds = diheds.shape[1]
per_diheds = np.zeros((diheds.shape[0], num_diheds * 2))
per_diheds[:, 0:num_diheds] = np.sin(diheds)
per_diheds[:, num_diheds:num_diheds * 2] = np.cos(diheds)
distances = distance.compute_distances(traj,
Dipeptide.PAIRS,
periodic=True,
opt=False)
distances = np.log(distances)
# Save
traj_list.append(distances)
traj_list_angles.append(diheds)
# traj_list.append(per_diheds)
self.traj_list = traj_list
self.traj_list_angles = traj_list_angles
def prepare_trajectory(self, trajectory):
"""Prepare a trajectory for distance calculations based on the contact map.
Each frame in the trajectory will be represented by a vector where
each entries represents the distance between two residues in the structure.
Depending on what contacts you pick to use, this can be a 'native biased'
picture or not.
Paramters
---------
trajectory : mdtraj.Trajectory
The trajectory to prepare
Returns
-------
pairwise_distances : ndarray
1D array of various residue-residue distances
"""
xyzlist = trajectory.xyz
num_residues = trajectory.n_residues
num_atoms = trajectory.n_atoms
if self.contacts == 'all':
contacts = np.empty(((num_residues - 2) * (num_residues - 3) / 2, 2), dtype=np.int32)
p = 0
for (a, b) in itertools.combinations(range(num_residues), 2):
if max(a, b) > min(a, b) + 2:
contacts[p, :] = [a, b]
p += 1
assert p == len(contacts), 'Something went wrong generating "all"'
else:
num, width = self.contacts.shape
contacts = self.contacts
if not width == 2:
raise ValueError('contacts must be width 2')
if not (0 < len(np.unique(contacts[:, 0])) < num_residues):
raise ValueError('contacts should refer to zero-based indexing of the residues')
if not np.all(np.logical_and(0 <= np.unique(contacts), np.unique(contacts) < num_residues)):
raise ValueError('contacts should refer to zero-based indexing of the residues')
if self.scheme == 'ca':
atom_contacts = np.zeros_like(contacts)
residue_to_alpha = np.zeros(num_residues) # zero based indexing
for i in range(num_atoms):
if trajectory.topology.atom(i).name == 'CA':
residue = trajectory.topology.residue(i).index - 1
residue_to_alpha[residue] = i
# print 'contacts (residues)', contacts
# print 'residue_to_alpja', residue_to_alpha.shape
# print 'residue_to_alpja', residue_to_alpha
atom_contacts = residue_to_alpha[contacts]
# print 'atom_contacts', atom_contacts
output = distance.compute_distances(trajectory, atom_contacts)
elif self.scheme in ['closest', 'closest-heavy']:
if self.scheme == 'closest':
residue_membership = [[atom.index for atom in residue.atoms]
for residue in trajectory.topology.residues]
elif self.scheme == 'closest-heavy':
residue_membership = [[] for i in range(num_residues)]
for i in range(num_atoms):
residue = trajectory.topology.atom(i).residue.index - 1
if not trajectory.topology.atom(i).name.lstrip('0123456789').startswith('H'):
residue_membership[residue].append(i)
# print 'Residue Membership'
# print residue_membership
# for row in residue_membership:
# for col in row:
# print "%s-%s" % (trajectory['AtomNames'][col], trajectory['ResidueID'][col]),
# print
output = _contactcalc.residue_distances(xyzlist, residue_membership, contacts)
else:
raise ValueError('This is not supposed to happen!')
return np.double(output)
def test_generator():
pairs2 = itertools.combinations(range(N_ATOMS), 2)
a = compute_distances(ptraj, pairs)
b = compute_distances(ptraj, pairs2)
eq(a, b)
def compute_rdf(traj,
pairs,
r_range=None,
bin_width=0.005,
n_bins=None,
periodic=True,
opt=True):
"""Compute radial distribution functions for pairs in every frame.
Parameters
----------
traj : Trajectory
Trajectory to compute radial distribution function in.
pairs : array-like, shape=(n_pairs, 2), dtype=int
Each row gives the indices of two atoms.
r_range : array-like, shape=(2,), optional, default=(0.0, 1.0)
Minimum and maximum radii.
bin_width : float, optional, default=0.005
Width of the bins in nanometers.
n_bins : int, optional, default=None
The number of bins. If specified, this will override the `bin_width`
parameter.
periodic : bool, default=True
If `periodic` is True and the trajectory contains unitcell
information, we will compute distances under the minimum image
convention.
opt : bool, default=True
Use an optimized native library to compute the pair wise distances.
Returns
-------
r : np.ndarray, shape=(np.diff(r_range) / bin_width - 1), dtype=float
Radii values corresponding to the centers of the bins.
g_r : np.ndarray, shape=(np.diff(r_range) / bin_width - 1), dtype=float
Radial distribution function values at r.
See also
--------
Topology.select_pairs
"""
if r_range is None:
r_range = np.array([0.0, 1.0])
r_range = ensure_type(r_range,
dtype=np.float64,
ndim=1,
name='r_range',
shape=(2, ),
warn_on_cast=False)
if n_bins is not None:
n_bins = int(n_bins)
if n_bins <= 0:
raise ValueError('`n_bins` must be a positive integer')
else:
n_bins = int((r_range[1] - r_range[0]) / bin_width)
distances = compute_distances(traj, pairs, periodic=periodic, opt=opt)
g_r, edges = np.histogram(distances, range=r_range, bins=n_bins)
r = 0.5 * (edges[1:] + edges[:-1])
# Normalize by volume of the spherical shell.
# See discussion https://github.com/mdtraj/mdtraj/pull/724. There might be
# a less biased way to accomplish this. The conclusion was that this could
# be interesting to try, but is likely not hugely consequential. This method
# of doing the calculations matches the implementation in other packages like
# AmberTools' cpptraj and gromacs g_rdf.
V = (4 / 3) * np.pi * (np.power(edges[1:], 3) - np.power(edges[:-1], 3))
norm = len(pairs) * np.sum(1.0 / traj.unitcell_volumes) * V
g_r = g_r.astype(np.float64) / norm # From int64.
return r, g_r
def prepare_trajectory(self, trajectory):
ptraj = distance.compute_distances(trajectory, self.atom_pairs)
return np.double(ptraj)
def test_3p():
a = compute_distances(ptraj, pairs, periodic=True, opt=True)
b = compute_displacements(ptraj, pairs, periodic=True, opt=True)
eq(a, np.sqrt(np.sum(np.square(b), axis=2)))
atom_pairs = []
for this_lig in inds_N:
for lig_atom, rec_atom in zip(this_lig, ind_res[0]):
atom_pairs.append([lig_atom, rec_atom])
atom_pairs = np.array(atom_pairs)
print 'We are tracking distances between the following pairs of atoms:'
print atom_pairs
# comput and record distances between ligand and binding pocket ASP113
distances = []
for this_sim in simulations:
for this_atom_pair in atom_pairs:
this_lig = []
for this_traj in [this_sim]:
this_lig.extend(
md_dist.compute_distances(this_traj, [this_atom_pair]))
distances.append(this_lig)
dist_path = '/home/shenglan/TryMSMbuilder/output/ten_ligands/dist_to_binding'\
+'_s'+str(LOAD_STRIDE)+'.out'
pickle.dump(distances, open(dist_path, 'wb'))
# get N positions
sequences_all = []
for this_sim in simulations:
this_seq = util.featurize_RawPos(inds_N, [this_sim])
sequences_all.extend(this_seq)
seq_path = '/home/shenglan/TryMSMbuilder/output/ten_ligands/sequences' + '_s' + str(
LOAD_STRIDE) + '.out'
pickle.dump(sequences_all, open(seq_path, 'wb'))
ii = [atom.index for atom in this_res.atoms if atom.name in ['OD1']]
ind_res.append(ii)
atom_pairs = []
for this_lig in inds_N:
for lig_atom,rec_atom in zip(this_lig,ind_res[0]):
atom_pairs.append([lig_atom,rec_atom])
atom_pairs = np.array(atom_pairs)
print 'We are tracking distances between the following pairs of atoms:'
print atom_pairs
distances = []
for this_sim in simulations:
for this_atom_pair in atom_pairs:
this_lig = []
for this_traj in this_sim:
this_lig.extend(md_dist.compute_distances(this_traj,[this_atom_pair]))
distances.append(this_lig)
#sequences of coordinates of ligands
total_frames = 0
sequences_all = []
for this_sim in simulations:
if use_COM:
this_seq = util.featurize_RawPos(inds_all,this_sim,average = True)
else:
this_seq = util.featurize_RawPos(inds_N,this_sim)
total_frames = this_seq[0].shape[0]*10+total_frames
sequences_all.extend(this_seq)
res_ind = [[atom.index for atom in target_res.atoms if atom.name in ['OD1']]]
atom_pairs = []
for this_lig in inds_N:
for lig_atom,rec_atom in zip(this_lig,res_ind[0]):
atom_pairs.append([lig_atom,rec_atom])
atom_pairs = np.array(atom_pairs)
print 'We are tracking distances between the following pairs of atoms:'
print atom_pairs
distances = []
for this_sim in simulations:
this_dist = []
for this_traj in this_sim:
this_dist.extend(md_dist.compute_distances(this_traj,atom_pairs))
this_dist = np.array(this_dist)
distances.append(this_dist)
sequences_all = []
total_frames = 0
for this_sim in simulations:
if use_COM:
this_seq = util.featurize_RawPos(inds_all,this_sim,average = True)
else:
this_seq = util.featurize_RawPos(inds_N,this_sim)
total_frames = this_seq[0].shape[0]*10+total_frames
sequences_all.extend(this_seq)
print('the total number of conformations we are clustering is %d.' % total_frames)
# convert to pdb so I can view in vmd
ind_res.append(ii)
atom_pairs = []
for this_lig in inds_N:
for lig_atom,rec_atom in zip(this_lig,ind_res[0]):
atom_pairs.append([lig_atom,rec_atom])
atom_pairs = np.array(atom_pairs)
print 'We are tracking distances between the following pairs of atoms:'
print atom_pairs
# comput and record distances between ligand and binding pocket ASP113
distances = []
for this_sim in simulations:
for this_atom_pair in atom_pairs:
this_lig = []
for this_traj in [this_sim]:
this_lig.extend(md_dist.compute_distances(this_traj,[this_atom_pair]))
distances.append(this_lig)
dist_path = '/home/shenglan/TryMSMbuilder/output/ten_ligands/dist_to_binding'\
+'_s'+str(LOAD_STRIDE)+'.out'
pickle.dump(distances,open(dist_path,'wb'))
# get N positions
sequences_all = []
for this_sim in simulations:
this_seq = util.featurize_RawPos(inds_N,[this_sim])
sequences_all.extend(this_seq)
seq_path = '/home/shenglan/TryMSMbuilder/output/ten_ligands/sequences'+'_s'+str(LOAD_STRIDE)+'.out'
pickle.dump(sequences_all,open(seq_path,'wb'))
clustering = KCenters(n_clusters = N_CLUSTER)
def test_0():
a = compute_distances(ptraj, pairs, periodic=False, opt=True)
b = compute_distances(ptraj, pairs, periodic=False, opt=False)
eq(a, b)
def test_generator():
pairs2 = itertools.combinations(range(N_ATOMS), 2)
a = compute_distances(ptraj, pairs)
b = compute_distances(ptraj, pairs2)
eq(a, b)
def test_0p():
a = compute_distances(ptraj, pairs, periodic=True, opt=True)
b = compute_distances(ptraj, pairs, periodic=True, opt=False)
eq(a, b, decimal=3)
def test_0():
a = compute_distances(ptraj, pairs, periodic=False, opt=True)
b = compute_distances(ptraj, pairs, periodic=False, opt=False)
eq(a, b)
def test_3p():
a = compute_distances(ptraj, pairs, periodic=True, opt=True)
b = compute_displacements(ptraj, pairs, periodic=True, opt=True)
eq(a, np.sqrt(np.sum(np.square(b), axis=2)))
def test_0p():
a = compute_distances(ptraj, pairs, periodic=True, opt=True)
b = compute_distances(ptraj, pairs, periodic=True, opt=False)
eq(a, b, decimal=3)
