def dcd_fit(filename_dcd, filename_dcd_out, natom_total, serials,
filename_rmsd):
f_out = open(filename_rmsd, 'w')
# Coord1
pdb = PdbFile('16SCD.cg.pdb')
pdb.open_to_read()
ref_chains = pdb.read_all()
pdb.close()
#num_atom = 0
#for chain in ref_chains :
# for residue in chain.residues :
# num_atom += len(residue.atoms)
ref = zeros((natom_total, 3), dtype=float64, order='C')
i = 0
for chain in ref_chains:
for residue in chain.residues:
for atom in residue.atoms:
(ref[i][0], ref[i][1], ref[i][2]) = atom.xyz.get_as_tuple()
i += 1
mask = []
for i in range(natom_total):
# the serial ID starts from 1, thus i+1 is the serial ID
if i + 1 in serials:
mask.append(1)
else:
mask.append(0)
dcd = DcdFile(filename_dcd)
dcd.open_to_read()
dcd.read_header()
out_dcd = DcdFile(filename_dcd_out)
out_dcd.open_to_write()
out_dcd.set_header(dcd.get_header())
out_dcd.write_header()
#dcd.show_header()
k = 0
while dcd.has_more_data():
k += 1
coords_dcd = dcd.read_onestep_np()
rmsd = superimpose(ref.T, coords_dcd.T, mask)
f_out.write('{:8d} {:6.2f}\n'.format(k, rmsd))
out_dcd.write_onestep(coords_dcd)
dcd.close()
out_dcd.close()
def dcd_concatenate(filepaths):
n_frames = [] # To return the number of frames
# Number of input DCD files
num_dcd = len(filepaths) - 1
# New DCD file
f_out = DcdFile(filepaths[-1])
f_out.open_to_write()
# Count the total frame number
num_frame = 1
for i in range(0, num_dcd - 1):
f_in = DcdFile(filepaths[i])
f_in.open_to_read()
f_in.read_header()
num_frame += f_in.get_header().nset - 1
f_in.close()
# Get the total step number from final DCD file
f_in = DcdFile(filepaths[num_dcd - 1])
f_in.open_to_read()
f_in.read_header()
num_frame += f_in.get_header().nset - 1
num_step = f_in.get_header().nstep
f_in.close()
f_in = DcdFile(filepaths[0])
f_in.open_to_read()
f_in.read_header()
header = copy(f_in.get_header())
header.nset = num_frame
header.nstep = num_step
f_out.set_header(header)
f_out.write_header()
#print filepaths[0], f_in.get_header().nset
n_frames.append(f_in.get_header().nset)
while f_in.has_more_data():
f_out.write_onestep(f_in.read_onestep())
f_in.close()
for i in range(1, num_dcd):
f_in = DcdFile(filepaths[i])
f_in.open_to_read()
f_in.read_header()
f_in.skip_onestep() # skip the first step
#print filepaths[i], f_in.get_header().nset - 1
n_frames.append(f_in.get_header().nset - 1)
while f_in.has_more_data():
f_out.write_onestep(f_in.read_onestep())
f_in.close()
f_out.close()
return n_frames
def count_frame(path):
dcd = DcdFile(path)
dcd.open_to_read()
dcd.read_header()
icount = 0
while dcd.has_more_data():
try:
dcd.skip(1)
icount += 1
except EOFError:
icount += 1
print(
'There is another frame at the end, that is not written completely.'
)
break
except:
break
dcd.close()
return icount
dist_array = []
while dcd.has_more_data():
ref = dcd.read_onestep_np()
dcd.set_mark()
if dcd.has_more_data():
while dcd.has_more_data():
data = dcd.read_onestep_np()
dist_array.append(calcrmsd(ref.T, data.T))
dcd.go_mark()
else:
break
dcd.close()
print(len(dist_array))
#z = scipy.cluster.hierarchy.ward(dist) # This does not work for some reason
z = scipy.cluster.hierarchy.linkage(dist_array,
method='ward',
metric='euclidean')
## Other methods
#z = scipy.cluster.hierarchy.linkage(data, method='single', metric='euclidean')
#z = scipy.cluster.hierarchy.linkage(data, method='average', metric='euclidean')
f_out = open(prefix + '.dcd.cls.z', 'w')
for iz, zz in enumerate(z):
f_out.write("%5i %5i %5i %f %i\n" % (iz + 1, zz[0], zz[1], zz[2], zz[3]))
f_out.close()
num_data += 1
data = f_dcd.read_onestep()
idx = 0
for iatom, atom in enumerate(data) :
if not fil[iatom]:
continue
for xyz in atom :
data_ave[idx] += xyz
idx += 1
for i in range(len(data_ave)) :
data_ave[i] /= float(num_data)
f_ave.write('%12.5f\n' % data_ave[i])
f_ave.close()
f_dcd.close()
f_dcd.open_to_read()
f_dcd.read_header()
# Calculate principal coordinate
while f_dcd.has_more_data() :
data = f_dcd.read_onestep()
for v in ev :
x = 0.0
idx = 0
for iatom, atom in enumerate(data) :
if not fil[iatom]:
continue
for xyz in atom :
x += (xyz - data_ave[idx]) * v[idx]
# read and write
#imodel = 0
#i_org = 0
while dcd.has_more_data():
if not ts.has_more_data():
print('Not enough data in .ts file (1)')
sys.exit(2)
tsdata, lines = ts.read_onestep()
# skip step=1
if tsdata[0][ts.head_col.step] == 1:
if not ts.has_more_data():
print('Not enough data in .ts file (2)')
sys.exit(2)
tsdata, lines = ts.read_onestep()
if tsdata[0][ts.head_col.e_bridge] == '0.0000000':
dcd_out.write_onestep(dcd.read_onestep())
#imodel += 1
else:
dcd.skip_onestep()
#i_org += 1
ts.close()
dcd.close()
dcd_out.close()
# Number of input DCD files
num_dcd = len(sys.argv) - 2
# New DCD file
f_out = DcdFile(sys.argv[-1])
f_out.open_to_write()
# Count the total frame number
num_frame = 1
for i in range(1, num_dcd):
f_in = DcdFile(sys.argv[i])
f_in.open_to_read()
f_in.read_header()
num_frame += f_in.get_header().nset
f_in.close()
# Get the total step number from final DCD file
f_in = DcdFile(sys.argv[num_dcd])
f_in.open_to_read()
f_in.read_header()
num_frame += f_in.get_header().nset
num_step = f_in.get_header().nstep
f_in.close()
f_in = DcdFile(sys.argv[1])
f_in.open_to_read()
f_in.read_header()
header = copy(f_in.get_header())
header.nset = num_frame
header.nstep = num_step
for i in range(1,ncls+1):
f_out.write('#cluster %i\n' % (i,))
for iframe, f in enumerate(fcls):
if f == i:
f_out.write("%i %i\n" % (iframe, iframe*nskip))
f_out.write('\n\n')
f_out.close()
############################# Center
dcd_ref = DcdFile(dcd_ref_filepath)
dcd_ref.open_to_read()
dcd_ref.read_header()
data_ref = dcd_ref.read_onestep_np()
dcd_ref.close()
dcd = DcdFile(dcd_filepath)
dcd.open_to_read()
dcd.read_header()
nmp = dcd._header.nmp_real
cls_centroids = []
for icls in range(ncls):
cls_centroids.append( np.zeros((nmp,3)) )
k = 0
while dcd.has_more_data():
data = dcd.read_onestep_np()
rmsd = superimpose(data_ref.T, data.T)
for i in range(NATOM_TOTAL):
if i in (309-1, 477-1, 519-1):
mask.append(1)
else:
mask.append(0)
dcd = DcdFile('dcd/cM0.0300.dcd')
dcd.open_to_read()
dcd.read_header()
out_dcd = DcdFile('cM0.0300.fit665.dcd')
out_dcd.open_to_write()
out_dcd.set_header(dcd.get_header())
out_dcd.write_header()
#dcd.show_header()
k = 0
while dcd.has_more_data() :
k += 1
coords_dcd = dcd.read_onestep_np()
rmsd = superimpose(ref.T, coords_dcd.T, mask)
print(k, rmsd)
out_dcd.write_onestep(coords_dcd)
dcd.close()
out_dcd.close()
nstep_dcd = 0
while in_dcd.has_more_data():
out_dcd.write_onestep( in_dcd.read_onestep() )
nstep_dcd += 1
nstep_ts = 0
while in_ts.has_more_data():
tsdata, tslines = in_ts.read_onestep()
out_ts.write_onestep( tslines )
nstep_ts += 1
if i_dir == 0:
print(('TS in %s: #frame = %i' % (d, nstep_ts)))
print(('DCD in %s: #frame = %i' % (d, nstep_dcd)))
else:
print(('TS in %s: #frame = %i (the first step is eliminated.)' % (d, nstep_ts)))
print(('DCD in %s: #frame = %i (the first step is eliminated.)' % (d, nstep_dcd)))
if nstep_ts != nstep_dcd:
print( 'Warning: nstep_ts != nstep_dcd' )
nstep_total_ts += nstep_ts
nstep_total_dcd += nstep_dcd
in_ts.close()
in_dcd.close()
print(('Total # of frames in TS output to %s: %i' % (dir_out, nstep_total_ts)))
print(('Total # of frames in DCD output to %s: %i' % (dir_out, nstep_total_dcd)))
