def is_hbonded2(mybgf, atom1, atom2, d_crit=2.2):
"""
returns True if two atoms are h-bonded by a geometric definition.
This definition is introduced in Grishina, N., & Buch, V. (2004). J. Chem. Phys., 120(11), 5217.
20160608: Works only for O atoms.
"""
if not "O" in atom1.ffType or not "O" in atom2.ffType:
return False
d = np.array([atom1.x, atom1.y, atom1.z])
a = np.array([atom2.x, atom2.y, atom2.z])
if mybgf.CRYSTX:
pbc = mybgf.CRYSTX[:3]
else:
pbc = 0
# A1-H ... A2
for ano in atom1.CONECT:
atom = mybgf.getAtom(ano) # this should be an H atom
x = np.array([atom.x, atom.y, atom.z])
if pbc:
dist = nu.pbc_dist(x, a, pbc)
else:
dist = nu.dist(x, a)
if 1e-5 < dist < d_crit:
return x
# A1 ... H-A2
for ano in atom2.CONECT:
atom = mybgf.getAtom(ano) # this should be an H atom
x = np.array([atom.x, atom.y, atom.z])
if pbc:
dist = nu.pbc_dist(x, a, pbc)
else:
dist = nu.dist(x, a)
if 1e-5 < dist < d_crit:
return x
return []
def countWaterCNT(bgf_file,
trj_file,
d_crit=3.5,
selection='',
nsample=0,
silent=False):
### init
# constants
deg_109p47 = math.radians(109.47)
# variables
timestep = 0
l_timestep = []
line = []
n_header = 0
avg_angles = []
avg_diheds = []
bin = np.arange(0.0, 181.0, 1.0)
t1 = 0
t2 = 0
# clock
myBGF = bgf.BgfFile(bgf_file)
myTRJ = open(trj_file)
myTRJ.seek(0)
myDAT = open(bgf_file[:-4] + ".AOP.dat", "w")
myDAT.write("#timestep\tAOP\tF4\n")
# how many steps to go?
#mytrj = lt.lammpstrj(trj_file)
mytrj = Trj(trj_file)
l_timestep = mytrj.load()
n_timestep = len(l_timestep)
l_timestep.sort()
requested_timesteps = l_timestep[-nsample:]
print(
"Using neighboring water distance criteria %4.1f A (should be consistent with the coordination number)"
% d_crit)
print("The trajectory contains " + str(n_timestep) + " timesteps.")
# Find header of the trajectory file
while 1:
templine = myTRJ.readline()
line.append(templine.strip('\n').strip('ITEM: '))
n_header += 1
if "ITEM: ATOMS" in templine:
break
# INITIAL trajectory information
timestep = int(line[1])
natoms = int(line[3])
boxsize = [
line[5].split(' ')[0], line[5].split(' ')[1], line[6].split(' ')[0],
line[6].split(' ')[1], line[7].split(' ')[0], line[7].split(' ')[1]
]
boxsize = [float(i) for i in boxsize]
keywords = line[8].strip('ATOMS ')
# for every shot in the trajectory file update BGF and manipulate
dumpatom = get_line(trj_file)
processed_step = 0
t1 = t2 = 0
elapsed_time = 0
while 1:
### Show progress
t1 = time.time()
remaining_time = elapsed_time * (len(requested_timesteps) -
processed_step)
sys.stdout.write('\r' + "Reading timestep.. " + str(timestep) +
" (Remaining time: " +
"{0:4.1f}".format(remaining_time) + " seconds, " +
"{0:4.1f} minutes".format(remaining_time / 60) + ")")
sys.stdout.flush()
### Read
try:
chunk = [next(dumpatom) for i in range(natoms + n_header)]
except StopIteration:
break
timestep = int(chunk[1])
natoms = int(chunk[3])
boxsize = [
chunk[5].split(' ')[0], chunk[5].split(' ')[1],
chunk[6].split(' ')[0], chunk[6].split(' ')[1],
chunk[7].split(' ')[0], chunk[7].split(' ')[1]
]
boxsize = [float(i) for i in boxsize]
boxsize = [(boxsize[1] - boxsize[0]), (boxsize[3] - boxsize[2]),
(boxsize[5] - boxsize[4])]
keywords = chunk[8].split('ATOMS ')[1].strip('\n').split(' ')
if not timestep in requested_timesteps:
continue
### update myBGF with trajectory information ###
natom_bgf = len(myBGF.a) # number of atoms in BGF file
if not natom_bgf == natoms:
nu.die(
"Number of atoms in trajectory file does not match with BGF file."
)
mode = ""
if 'xs' in keywords or 'ys' in keywords or 'zs' in keywords:
mode = 'scaled'
elif 'x' in keywords or 'y' in keywords or 'z' in keywords:
mode = 'normal'
elif 'xu' in keywords or 'yu' in keywords or 'zu' in keywords:
mode = 'unwrapped'
# actual coordinate
coordinfo = chunk[9:]
# assume that coordinfo is similar to ['id', 'type', 'xs', 'ys', 'zs', 'ix', 'iy', 'iz']
for atomline in coordinfo:
atomcoord = atomline.split(' ')
atom = myBGF.getAtom(int(atomcoord[0]))
if mode == 'scaled':
atom.x = float(atomcoord[2]) * boxsize[0]
atom.y = float(atomcoord[3]) * boxsize[1]
atom.z = float(atomcoord[4]) * boxsize[2]
elif mode == 'unwrapped':
atom.x = float(atomcoord[2])
atom.y = float(atomcoord[3])
atom.z = float(atomcoord[4])
elif mode == 'normal':
try:
ix_index = keywords.index('ix')
iy_index = keywords.index('iy')
iz_index = keywords.index('iz')
except ValueError:
nu.warn(
"No image information no the trajectory file. Will be treated as unwrapped."
)
atom.x = float(atomcoord[2])
atom.y = float(atomcoord[3])
atom.z = float(atomcoord[4])
else:
atom.x = (int(atomcoord[ix_index]) * boxsize[0]) + float(
atomcoord[2])
atom.y = (int(atomcoord[iy_index]) * boxsize[1]) + float(
atomcoord[3])
atom.z = (int(atomcoord[iz_index]) * boxsize[2]) + float(
atomcoord[4])
try:
for i in range(0, 3):
myBGF.CRYSTX[i] = boxsize[i]
except:
pass
# apply periodic condition
myBGF = bgftools.periodicMoleculeSort(myBGF, myBGF.CRYSTX, silent=True)
### myBGF update complete! ###
# get water OW
O = []
anos_O = []
for atom in myBGF.a:
if selection:
if "O" in atom.ffType and eval(selection):
O.append(atom)
anos_O.append(atom.aNo)
else:
if "O" in atom.ffType:
O.append(atom)
anos_O.append(atom.aNo)
if not len(O):
nu.die("There are no O atoms which satisfies %s!" % selection)
### calculate AOP and F4
sys.stdout.write('AOP..... ')
sys.stdout.flush()
coords = []
anos = []
for atom in O:
coords.append([atom.x, atom.y, atom.z])
anos.append(atom.aNo)
pbc = np.array(myBGF.CRYSTX[:3])
coords = np.array(coords)
tree = pkdtree.PeriodicKDTree(
pbc, coords) # KDtree for distance calculation
# analyze local water structures for a timestep
n_neighbors = []
angles = []
diheds = []
for atom in O:
# local variables
# find neighbors
neighbors = []
# list of O atoms near centerO
centerO = np.array([atom.x, atom.y, atom.z])
d, ndx = tree.query(centerO, k=6)
index = np.where((d >= 1e-4) & (d <= d_crit)) # discard self
for i in ndx[index]:
neighbors.append(myBGF.getAtom(anos[i]))
# calculate O1-O-O2 angle
for i, j in itertools.combinations(neighbors, 2):
if not "O" in i.ffType or not "O" in j.ffType:
nu.die("Wrong atom found.")
x = [i.x, i.y, i.z]
y = [j.x, j.y, j.z]
pbc_dist = nu.pbc_dist(x, y, pbc)
dist = nu.dist(x, y)
if abs(pbc_dist - dist) > 0.1:
continue
# discard atoms over pbc boundaries
# angle
angle = nu.angle(x, centerO, y, radians=False)
angles.append(angle)
n_neighbors.append(
len(neighbors)) # number of neighboring water molecules
# calculate dihedral
for i in neighbors:
# find aNos of H located farthest among all combinations
hO_anos = atom.CONECT
hi_anos = i.CONECT
max_dist_pair = []
max_dist = 0.0
for k, l in itertools.product(hO_anos, hi_anos):
h1 = myBGF.getAtom(k) # H atom connected with atom
h2 = myBGF.getAtom(l) # H atom connected with i
dist = bgf.distance(h1, h2)
if dist > max_dist:
max_dist = dist
max_dist_pair = [
h1, h2
] # now we have two H atoms in maximum distance connected with atom and i
# dihedral angle
phi = bgf.dihedral(max_dist_pair[0],
atom,
i,
max_dist_pair[1],
radians=False)
diheds.append(phi)
hist_angle, _ = np.histogram(angles, bins=bin, normed=True)
hist_dihed, _ = np.histogram(diheds, bins=bin, normed=True)
avg_angles.append(hist_angle)
avg_diheds.append(hist_dihed)
sys.stdout.write(
'Done ')
sys.stdout.flush()
# write output
processed_step += 1
t2 = time.time() # time mark
elapsed_time = t2 - t1
avg_angles = np.mean(np.array(avg_angles), axis=0)
avg_diheds = np.mean(np.array(avg_diheds), axis=0)
#print(avg_diheds)
myDAT.write("#angles population\n")
for index, i in enumerate(avg_angles):
myDAT.write("%8.2f %8.5f\n" % (_[index], avg_angles[index]))
myDAT.write("\n\n")
myDAT.write("#diheds population\n")
for index, i in enumerate(avg_diheds):
myDAT.write("%8.2f %8.5f\n" % (_[index], avg_diheds[index]))
myDAT.close()
print('')
return 1
def getHBond(bgf_file, trj_file, selection='', silent=False):
# init
timestep = 0
l_timestep = []
line = []
n_header = 0
t1 = 0
t2 = 0
# clock
hbond_dat = dict()
d_crit = 3.5
a_crit = 30.0
myBGF = bgf.BgfFile(bgf_file)
myTRJ = open(trj_file)
myTRJ.seek(0)
myPickle_file = bgf_file[:-4] + ".hbond.pickle"
myDAT = open(bgf_file[:-4] + ".hbond.count.dat", "w")
myDAT.write(str(sys.argv) + "\n")
# how many steps to go?
wc_trj_file = popen("grep TIMESTEP " + trj_file + " | wc -l ").read()
n_timestep = int(wc_trj_file.split()[0])
print("The trajectory contains " + str(n_timestep) + " timesteps.")
# Find header of the trajectory file
while 1:
templine = myTRJ.readline()
line.append(templine.strip('\n').strip('ITEM: '))
n_header += 1
if "ITEM: ATOMS" in templine:
break
# INITIAL trajectory information
timestep = int(line[1])
natoms = int(line[3])
boxsize = [
line[5].split(' ')[0], line[5].split(' ')[1], line[6].split(' ')[0],
line[6].split(' ')[1], line[7].split(' ')[0], line[7].split(' ')[1]
]
boxsize = [float(i) for i in boxsize]
keywords = line[8].strip('ATOMS ')
# for every shot in the trajectory file update BGF and manipulate
dumpatom = get_line(trj_file)
processed_step = 0
t1 = t2 = 0
elapsed_time = 0
while 1:
### Show progress
t1 = time.time()
remaining_time = elapsed_time * (n_timestep - processed_step)
sys.stdout.write('\r' + "Reading timestep.. " + str(timestep) +
" (Remaining time: " +
"{0:4.1f}".format(remaining_time) + " seconds, " +
"{0:4.1f} minutes".format(remaining_time / 60) + ")")
sys.stdout.flush()
### Read
try:
chunk = [next(dumpatom) for i in range(natoms + n_header)]
except StopIteration:
break
timestep = int(chunk[1])
natoms = int(chunk[3])
boxsize = [
chunk[5].split(' ')[0], chunk[5].split(' ')[1],
chunk[6].split(' ')[0], chunk[6].split(' ')[1],
chunk[7].split(' ')[0], chunk[7].split(' ')[1]
]
boxsize = [float(i) for i in boxsize]
boxsize = [(boxsize[1] - boxsize[0]), (boxsize[3] - boxsize[2]),
(boxsize[5] - boxsize[4])]
keywords = chunk[8].split('ATOMS ')[1].strip('\n').split(' ')
### update myBGF with trajectory information ###
natom_bgf = len(myBGF.a) # number of atoms in BGF file
if not natom_bgf == natoms:
nu.die(
"Number of atoms in trajectory file does not match with BGF file."
)
mode = ""
if 'xs' in keywords or 'ys' in keywords or 'zs' in keywords:
mode = 'scaled'
elif 'x' in keywords or 'y' in keywords or 'z' in keywords:
mode = 'normal'
elif 'xu' in keywords or 'yu' in keywords or 'zu' in keywords:
mode = 'unwrapped'
# actual coordinate
coordinfo = chunk[9:]
# assume that coordinfo is similar to ['id', 'type', 'xs', 'ys', 'zs', 'ix', 'iy', 'iz']
for atomline in coordinfo:
atomcoord = atomline.split(' ')
atom = myBGF.getAtom(int(atomcoord[0]))
if mode == 'scaled':
atom.x = float(atomcoord[2]) * boxsize[0]
atom.y = float(atomcoord[3]) * boxsize[1]
atom.z = float(atomcoord[4]) * boxsize[2]
elif mode == 'unwrapped':
atom.x = float(atomcoord[2])
atom.y = float(atomcoord[3])
atom.z = float(atomcoord[4])
elif mode == 'normal':
try:
ix_index = keywords.index('ix')
iy_index = keywords.index('iy')
iz_index = keywords.index('iz')
except ValueError:
nu.warn(
"No image information no the trajectory file. Will be treated as unwrapped."
)
atom.x = float(atomcoord[2])
atom.y = float(atomcoord[3])
atom.z = float(atomcoord[4])
else:
atom.x = (int(atomcoord[ix_index]) * boxsize[0]) + float(
atomcoord[2])
atom.y = (int(atomcoord[iy_index]) * boxsize[1]) + float(
atomcoord[3])
atom.z = (int(atomcoord[iz_index]) * boxsize[2]) + float(
atomcoord[4])
try:
for i in range(0, 3):
myBGF.CRYSTX[i] = boxsize[i]
except:
pass
#nu.warn("Crystal information error: is this file not periodic?")
pbc = myBGF.CRYSTX[:3]
# apply periodic condition
myBGF = bgftools.periodicMoleculeSort(myBGF, myBGF.CRYSTX, silent=True)
### myBGF update complete! ###
# get donor and acceptor
# donors: O in ACT
A = []
D = []
for atom in myBGF.a:
if selection:
if "O" in atom.ffType and eval(selection):
A.append(atom)
if "O" in atom.ffType and eval(selection):
D.append(atom)
else:
if "O" in atom.ffType:
A.append(atom)
if "O" in atom.ffType:
D.append(atom)
if not len(A) or not len(D):
nu.die(
"There are no atoms which can make H_bond (especially O atoms)!"
)
### find hydrogen bonds
# for all pairs of OC-HW
sys.stdout.write('Hbonds.. ')
sys.stdout.flush()
hbonds = []
donors = []
acceptors = []
hydrogens = []
angles = []
distances = []
for d_atom in D:
d = np.array([d_atom.x, d_atom.y, d_atom.z])
for a_atom in A:
a = np.array([a_atom.x, a_atom.y, a_atom.z])
dist = nu.dist(d, a)
#dist = bgf.distance(d_atom, a_atom)
if 0.001 < dist < d_crit:
# check H angle
for ano in d_atom.CONECT:
h_atom = myBGF.getAtom(ano)
h = np.array([h_atom.x, h_atom.y, h_atom.z])
u = h - d
v = a - d
theta = np.dot(u, v) / norm(u) / norm(v)
theta = np.degrees(arccos(theta))
#theta = bgf.angle(h_atom, d_atom, a_atom, radians=False)
if theta < a_crit:
hbonds.append([d_atom.aNo, a_atom.aNo])
donors.append(d_atom.aNo)
acceptors.append(a_atom.aNo)
hydrogens.append(h_atom.aNo)
distances.append(dist)
angles.append(theta)
#n_hbond += 1
#print("%s %s" % (d_atom.rNo, a_atom.rNo))
hbond_dat[timestep] = hbonds
sys.stdout.write(
'Done ')
sys.stdout.flush()
print("")
for i in zip(donors, acceptors, hydrogens, distances, angles):
print i
sys.exit(0)
myDAT.write("%d %d\n" % (timestep, len(hbonds)))
# write output
#myBGF.saveBGF(bgf_file[:-4] + "." + str(timestep) + ".bgf")
#myBGF2.saveBGF(bgf_file[:-4] + "." + str(timestep) + ".bgf")
t2 = time.time() # time mark
elapsed_time = t2 - t1
pickle_f = open(myPickle_file, 'w')
pickle.dump(hbond_dat, pickle_f)
print('')
myDAT.close()
return 1
for atom in tqdm.tqdm(moslayer.a, ncols=120, desc='Assigning ffTypes'):
if "S" in atom.ffType:
if atom.z > avg_mo_z:
atom.ffType = "S_3a"
else:
atom.ffType = "S_3b"
# remove infinite boundary
n_del = 0
for atom in tqdm.tqdm(moslayer.a, ncols=120, desc='Removing pbc bonds'):
a = [atom.x, atom.y, atom.z]
connected_ano = atom.CONECT
for ano in connected_ano:
atom2 = moslayer.getAtom(ano)
a2 = [atom2.x, atom2.y, atom2.z]
dist = nu.dist(a, a2)
pbc_dist = nu.pbc_dist(a, a2, moslayer.CRYSTX[:3])
if dist != pbc_dist:
moslayer.disconnect(moslayer.a2i[atom.aNo],
moslayer.a2i[atom2.aNo])
n_del += 1
print("%d bonds are disconnected." % n_del)
# remove infinite boundary
n_del = 0
for atom in tqdm.tqdm(moslayer.a, ncols=120, desc='Removing pbc bonds 2'):
a = [atom.x, atom.y, atom.z]
connected_ano = atom.CONECT
for ano in connected_ano:
atom2 = moslayer.getAtom(ano)
a2 = [atom2.x, atom2.y, atom2.z]
def analyze(bgf_file, trj_file, ff_file='', name='', selection='', d_crit=3.5):
'''analyze something within a timestep in a series of lammps trajectory.
'''
# variables
result_angles = dict()
result_diheds = dict()
print(
"Using neighboring water distance criteria %4.1f A (should be consistent with the coordination number)"
% d_crit)
# inner functions
def get_line(file):
with open(file, 'r') as f:
for line in f:
yield line
# 1. Load BGF
mybgf = bgf.BgfFile(bgf_file)
N_BGF_ATOMS = len(mybgf.a)
# 2. Read LAMMPS Trajectory
#timesteps, N_HEADER, N_ATOMS = lt.getTrjInfo(trj_file)
mytrj = lt.lammpstrj(trj_file)
timesteps = mytrj.load()
N_HEADER = mytrj.nheader
N_ATOMS = mytrj.natoms[0]
N_BUFFER = N_HEADER + N_ATOMS
if N_BGF_ATOMS != N_ATOMS:
nu.die(
"Number of atoms in trajectory file does not match with BGF file.")
# 3. Determine dump style
dump_keywords = mytrj.dumpstyle
yes_scale = False
if 'xs' in dump_keywords:
yes_scale = True
# 4. Update coordinates from the snapshot
dump = get_line(trj_file)
for t in tqdm.tqdm(timesteps, ncols=120, desc="Analyzing"):
chunk = [next(dump) for i in range(N_BUFFER)]
t = int(chunk[1])
mybgf.CRYSTX = mytrj.pbc[t] + [90.0, 90.0, 90.0]
coords = chunk[9:]
for c in coords:
c = c.split(' ')
atom = mybgf.getAtom(int(c[0]))
if yes_scale:
atom.x = float(c[2]) * pbc[0]
atom.y = float(c[3]) * pbc[1]
atom.z = float(c[4]) * pbc[2]
else:
atom.x = float(c[2])
atom.y = float(c[3])
atom.z = float(c[4])
mybgf = bt.periodicMoleculeSort(mybgf,
mybgf.CRYSTX,
ff_file=ff_file,
silent=True)
# Calculate Angles and Dihedrals for Water Structure
O = []
for atom in mybgf.a:
if selection:
if "O" in atom.ffType and eval(selection):
O.append(atom)
else:
if "O" in atom.ffType:
O.append(atom)
if not len(O):
nu.warn("There are no O atoms which satisfies %s!" % selection)
coords = []
anos = []
for atom in O:
coords.append([atom.x, atom.y, atom.z])
anos.append(atom.aNo)
coords = np.array(coords)
#tree = pkdtree.PeriodicKDTree(mytrj.pbc[t], coords) # KDTree with pbc
tree = scipy.spatial.KDTree(
coords, leafsize=len(coords) +
1) # REMARK: normal KDTree -- do not consider over pbc
angles = []
diheds = []
for atom in O:
# Find neighbors
coord = np.array([atom.x, atom.y, atom.z])
neighbor_O = []
# list of O atoms near centerO
d, ndx = tree.query(coord, k=5)
index = np.where((d >= 1e-4) & (d <= d_crit)) # discard self
for i in ndx[index]:
neighbor_O.append(mybgf.getAtom(anos[i]))
# Angles
for i, j in itertools.combinations(neighbor_O, 2):
x = [i.x, i.y, i.z]
y = [j.x, j.y, j.z]
angle = nu.angle(x, coord, y, radians=False)
angles.append([angle, i.z, atom.z,
j.z]) # angles: result_angle data structure
# Dihedrals: farthest H-O...O-H for all neighboring O-O pairs
for atom2 in neighbor_O:
# are they have an hydrogen bond?
if bt.is_hbonded(mybgf, atom, atom2):
hO_anos = atom.CONECT # H connected to center O
hO2_anos = atom2.CONECT # H connected to neighboring O
max_dist_pair = []
max_dist = 0.0
for k, l in itertools.product(hO_anos, hO2_anos):
h1 = mybgf.getAtom(k)
h2 = mybgf.getAtom(l)
dist = nu.dist([h1.x, h1.y, h1.z], [h2.x, h2.y, h2.z])
if dist > max_dist:
max_dist = dist
max_dist_pair = [h1, h2]
phi = bgf.dihedral(max_dist_pair[0],
atom,
atom2,
max_dist_pair[1],
radians=False)
diheds.append([phi, atom.z,
atom2.z]) # diheds: result_dihed data structure
result_angles[t] = angles
result_diheds[t] = diheds
# Write-ups
if name:
out_file = name + '.aop'
else:
out_file = 'aop'
angle_out_file = out_file + ".angle.pickle"
dihed_out_file = out_file + ".dihed.pickle"
with open(angle_out_file, 'wb') as f:
pickle.dump(result_angles, f, protocol=pickle.HIGHEST_PROTOCOL)
print("Success to save the angle result to a pickle file %s" %
angle_out_file)
with open(dihed_out_file, 'wb') as f:
pickle.dump(result_diheds, f, protocol=pickle.HIGHEST_PROTOCOL)
print("Success to save the dihed result to a pickle file %s" %
dihed_out_file)
print("Done.")