def _parsePDBLines(atomgroup, lines, split, model, chain, subset,
altloc_torf, format='PDB'):
"""Returns an AtomGroup. See also :func:`.parsePDBStream()`.
:arg lines: PDB/PQR lines
:arg split: starting index for coordinate data lines"""
format = format.upper()
if format == 'PDB':
isPDB = True
else:
isPDB = False
if subset:
if subset == 'ca':
subset = set(('CA',))
elif subset in 'bb':
subset = flags.BACKBONE
only_subset = True
protein_resnames = flags.AMINOACIDS
else:
only_subset = False
if chain is None:
only_chains = False
else:
only_chains = True
onlycoords = False
n_atoms = atomgroup.numAtoms()
if n_atoms > 0:
asize = n_atoms
else:
asize = len(lines) - split
addcoords = False
if atomgroup.numCoordsets() > 0:
addcoords = True
alength = asize
coordinates = np.zeros((asize, 3), dtype=float)
atomnames = np.zeros(asize, dtype=ATOMIC_FIELDS['name'].dtype)
resnames = np.zeros(asize, dtype=ATOMIC_FIELDS['resname'].dtype)
resnums = np.zeros(asize, dtype=ATOMIC_FIELDS['resnum'].dtype)
chainids = np.zeros(asize, dtype=ATOMIC_FIELDS['chain'].dtype)
hetero = np.zeros(asize, dtype=bool)
termini = np.zeros(asize, dtype=bool)
altlocs = np.zeros(asize, dtype=ATOMIC_FIELDS['altloc'].dtype)
icodes = np.zeros(asize, dtype=ATOMIC_FIELDS['icode'].dtype)
serials = np.zeros(asize, dtype=ATOMIC_FIELDS['serial'].dtype)
charges = np.zeros(asize, dtype=ATOMIC_FIELDS['charge'].dtype)
if isPDB:
segnames = np.zeros(asize, dtype=ATOMIC_FIELDS['segment'].dtype)
elements = np.zeros(asize, dtype=ATOMIC_FIELDS['element'].dtype)
bfactors = np.zeros(asize, dtype=ATOMIC_FIELDS['beta'].dtype)
occupancies = np.zeros(asize, dtype=ATOMIC_FIELDS['occupancy'].dtype)
anisou = None
siguij = None
else:
radii = np.zeros(asize, dtype=ATOMIC_FIELDS['radius'].dtype)
asize = 2000 # increase array length by this much when needed
start = split
stop = len(lines)
nmodel = 0
# if a specific model is requested, skip lines until that one
if isPDB and model is not None and model != 1:
for i in range(split, len(lines)):
if lines[i][:5] == 'MODEL':
nmodel += 1
if model == nmodel:
start = i+1
stop = len(lines)
break
if nmodel != model:
raise PDBParseError('model {0} is not found'.format(model))
if isinstance(altloc_torf, str):
if altloc_torf.strip() != 'A':
LOGGER.info('Parsing alternate locations {0}.'
.format(altloc_torf))
which_altlocs = ' ' + ''.join(altloc_torf.split())
else:
which_altlocs = ' A'
altloc_torf = False
else:
which_altlocs = ' A'
altloc_torf = True
acount = 0
coordsets = None
altloc = defaultdict(list)
i = start
END = False
while i < stop:
line = lines[i]
if not isPDB:
fields = line.split()
if len(fields) == 10:
fields.insert(4, '')
elif len(fields) != 11:
LOGGER.warn('wrong number of fields for PQR format at line %d'%i)
i += 1
continue
if isPDB:
startswith = line[0:6].strip()
else:
startswith = fields[0]
if startswith == 'ATOM' or startswith == 'HETATM':
if isPDB:
atomname = line[12:16].strip()
resname = line[17:21].strip()
else:
atomname= fields[2]
resname = fields[3]
if only_subset:
if not (atomname in subset and resname in protein_resnames):
i += 1
continue
if isPDB:
chid = line[21]
else:
chid = fields[4]
if only_chains:
if not chid in chain:
i += 1
continue
if isPDB:
alt = line[16]
if alt not in which_altlocs:
altloc[alt].append((line, i))
i += 1
continue
else:
alt = ' '
try:
if isPDB:
coordinates[acount, 0] = line[30:38]
coordinates[acount, 1] = line[38:46]
coordinates[acount, 2] = line[46:54]
else:
coordinates[acount, 0] = fields[6]
coordinates[acount, 1] = fields[7]
coordinates[acount, 2] = fields[8]
except:
if acount >= n_atoms > 0:
if nmodel == 0:
raise ValueError(format + 'file and AtomGroup ag must '
'have same number of atoms')
LOGGER.warn('Discarding model {0}, which contains {1} more '
'atoms than first model does.'
.format(nmodel+1,acount-n_atoms+1))
acount = 0
nmodel += 1
coordinates = np.zeros((n_atoms, 3), dtype=float)
if isPDB:
while lines[i][:6] != 'ENDMDL':
i += 1
else:
raise PDBParseError('invalid or missing coordinate(s) at '
'line {0}'.format(i+1))
if onlycoords:
acount += 1
i += 1
continue
try:
serials[acount] = int(line[6:11]) if isPDB else int(fields[1])
except ValueError:
try:
serials[acount] = int(line[6:11], 16) if isPDB else int(fields[1], 16)
except ValueError:
LOGGER.warn('failed to parse serial number in line {0}'
.format(i))
serials[acount] = serials[acount-1]+1
altlocs[acount] = alt
atomnames[acount] = atomname
resnames[acount] = resname
chainids[acount] = chid
if isPDB:
resnums[acount] = line[22:26]
icodes[acount] = line[26]
else:
resnum = fields[5]
if resnum[-1].isalpha():
icode = resnum[-1]
else:
icode = ' '
resnums[acount] = resnum
icodes[acount] = icode
if isPDB:
try:
occupancies[acount] = line[54:60]
except:
LOGGER.warn('failed to parse occupancy at line {0}'
.format(i))
try:
bfactors[acount] = line[60:66]
except:
LOGGER.warn('failed to parse beta-factor at line {0}'
.format(i))
hetero[acount] = startswith[0] == 'H'
segnames[acount] = line[72:76]
elements[acount] = line[76:78]
try:
charges[acount] = int(line[79] + line[78])
except:
charges[acount] = 0
else:
try:
charges[acount] = fields[9]
except:
LOGGER.warn('failed to parse charge at line {0}'
.format(i))
try:
radii[acount] = fields[10]
except:
LOGGER.warn('failed to parse radius at line {0}'
.format(i))
acount += 1
if n_atoms == 0 and acount >= alength:
# if arrays are short extend them with zeros
alength += asize
coordinates = np.concatenate(
(coordinates, np.zeros((asize, 3), float)))
atomnames = np.concatenate((atomnames,
np.zeros(asize, ATOMIC_FIELDS['name'].dtype)))
resnames = np.concatenate((resnames,
np.zeros(asize, ATOMIC_FIELDS['resname'].dtype)))
resnums = np.concatenate((resnums,
np.zeros(asize, ATOMIC_FIELDS['resnum'].dtype)))
chainids = np.concatenate((chainids,
np.zeros(asize, ATOMIC_FIELDS['chain'].dtype)))
hetero = np.concatenate((hetero, np.zeros(asize, bool)))
termini = np.concatenate((termini, np.zeros(asize, bool)))
altlocs = np.concatenate((altlocs,
np.zeros(asize, ATOMIC_FIELDS['altloc'].dtype)))
icodes = np.concatenate((icodes,
np.zeros(asize, ATOMIC_FIELDS['icode'].dtype)))
serials = np.concatenate((serials,
np.zeros(asize, ATOMIC_FIELDS['serial'].dtype)))
if isPDB:
bfactors = np.concatenate((bfactors,
np.zeros(asize, ATOMIC_FIELDS['beta'].dtype)))
occupancies = np.concatenate((occupancies,
np.zeros(asize, ATOMIC_FIELDS['occupancy'].dtype)))
segnames = np.concatenate((segnames,
np.zeros(asize, ATOMIC_FIELDS['segment'].dtype)))
elements = np.concatenate((elements,
np.zeros(asize, ATOMIC_FIELDS['element'].dtype)))
if anisou is not None:
anisou = np.concatenate((anisou, np.zeros((asize, 6),
ATOMIC_FIELDS['anisou'].dtype)))
if siguij is not None:
siguij = np.concatenate((siguij, np.zeros((asize, 6),
ATOMIC_FIELDS['siguij'].dtype)))
else:
charges = np.concatenate((charges,
np.zeros(asize, ATOMIC_FIELDS['charge'].dtype)))
radii = np.concatenate((radii,
np.zeros(asize, ATOMIC_FIELDS['radius'].dtype)))
#elif startswith == 'END ' or startswith == 'CONECT':
# i += 1
# break
elif not onlycoords and (startswith == 'TER ' or
startswith.strip() == 'TER'):
termini[acount - 1] = True
elif startswith == 'ENDMDL' or startswith[:3] == 'END':
if acount == 0:
# If there is no atom record between ENDMDL & END skip to next
i += 1
continue
if model is not None:
i += 1
break
diff = stop - i - 1
END = diff < acount
if coordsets is not None:
END = END or nmodel >= coordsets.shape[0]
if onlycoords:
if acount < n_atoms:
LOGGER.warn('Discarding model {0}, which contains '
'{1} fewer atoms than the first model '
'does.'.format(nmodel+1, n_atoms-acount))
else:
coordsets[nmodel] = coordinates
nmodel += 1
acount = 0
if not END:
coordinates = coordsets[nmodel]
else:
if acount != n_atoms > 0:
raise ValueError('PDB file and AtomGroup ag must have '
'same number of atoms')
# this is where to decide if more coordsets should be expected
if END:
coordinates.resize((acount, 3), refcheck=False)
if addcoords:
atomgroup.addCoordset(coordinates)
else:
atomgroup._setCoords(coordinates)
else:
coordsets = np.zeros((int(diff//acount+1), acount, 3))
coordsets[0] = coordinates[:acount]
onlycoords = True
atomnames.resize(acount, refcheck=False)
resnames.resize(acount, refcheck=False)
resnums.resize(acount, refcheck=False)
chainids.resize(acount, refcheck=False)
hetero.resize(acount, refcheck=False)
termini.resize(acount, refcheck=False)
altlocs.resize(acount, refcheck=False)
icodes.resize(acount, refcheck=False)
serials.resize(acount, refcheck=False)
if not only_subset:
atomnames = np.char.strip(atomnames)
resnames = np.char.strip(resnames)
atomgroup.setNames(atomnames)
atomgroup.setResnames(resnames)
atomgroup.setResnums(resnums)
atomgroup.setChids(chainids)
atomgroup.setFlags('hetatm', hetero)
atomgroup.setFlags('pdbter', termini)
atomgroup.setAltlocs(altlocs)
atomgroup.setIcodes(np.char.strip(icodes))
atomgroup.setSerials(serials)
if isPDB:
bfactors.resize(acount, refcheck=False)
occupancies.resize(acount, refcheck=False)
segnames.resize(acount, refcheck=False)
elements.resize(acount, refcheck=False)
atomgroup.setBetas(bfactors)
atomgroup.setOccupancies(occupancies)
atomgroup.setSegnames(np.char.strip(segnames))
atomgroup.setElements(np.char.strip(elements))
from prody.utilities.misctools import getMasses
atomgroup.setMasses(getMasses(np.char.strip(elements)))
if anisou is not None:
anisou.resize((acount, 6), refcheck=False)
atomgroup.setAnisous(anisou / 10000)
if siguij is not None:
siguij.resize((acount, 6), refcheck=False)
atomgroup.setAnistds(siguij / 10000)
else:
charges.resize(acount, refcheck=False)
radii.resize(acount, refcheck=False)
atomgroup.setCharges(charges)
atomgroup.setRadii(radii)
nmodel += 1
n_atoms = acount
acount = 0
coordinates = np.zeros((n_atoms, 3), dtype=float)
if altloc and altloc_torf:
_evalAltlocs(atomgroup, altloc, chainids, resnums,
resnames, atomnames)
altloc = defaultdict(list)
if END:
break
elif isPDB and startswith == 'ANISOU':
if anisou is None:
anisou = True
anisou = np.zeros((alength, 6),
dtype=ATOMIC_FIELDS['anisou'].dtype)
try:
index = acount - 1
anisou[index, 0] = line[28:35]
anisou[index, 1] = line[35:42]
anisou[index, 2] = line[43:49]
anisou[index, 3] = line[49:56]
anisou[index, 4] = line[56:63]
anisou[index, 5] = line[63:70]
except:
LOGGER.warn('failed to parse anisotropic temperature '
'factors at line {0}'.format(i))
elif isPDB and startswith =='SIGUIJ':
if siguij is None:
siguij = np.zeros((alength, 6),
dtype=ATOMIC_FIELDS['siguij'].dtype)
try:
index = acount - 1
siguij[index, 0] = line[28:35]
siguij[index, 1] = line[35:42]
siguij[index, 2] = line[43:49]
siguij[index, 3] = line[49:56]
siguij[index, 4] = line[56:63]
siguij[index, 5] = line[63:70]
except:
LOGGER.warn('failed to parse standard deviations of '
'anisotropic temperature factors at line {0}'.format(i))
elif startswith =='SIGATM':
pass
i += 1
if onlycoords:
if acount == atomgroup.numAtoms():
coordsets[nmodel] = coordinates
nmodel += 1
del coordinates
coordsets.resize((nmodel, atomgroup.numAtoms(), 3), refcheck=False)
if addcoords:
atomgroup.addCoordset(coordsets)
else:
atomgroup._setCoords(coordsets)
elif not END:
# this means last line was an ATOM line, so atomgroup is not decorated
coordinates.resize((acount, 3), refcheck=False)
if addcoords:
atomgroup.addCoordset(coordinates)
else:
atomgroup._setCoords(coordinates)
atomnames.resize(acount, refcheck=False)
resnames.resize(acount, refcheck=False)
resnums.resize(acount, refcheck=False)
chainids.resize(acount, refcheck=False)
hetero.resize(acount, refcheck=False)
termini.resize(acount, refcheck=False)
altlocs.resize(acount, refcheck=False)
icodes.resize(acount, refcheck=False)
serials.resize(acount, refcheck=False)
if not only_subset:
atomnames = np.char.strip(atomnames)
resnames = np.char.strip(resnames)
atomgroup.setNames(atomnames)
atomgroup.setResnames(resnames)
atomgroup.setResnums(resnums)
atomgroup.setChids(chainids)
atomgroup.setFlags('hetatm', hetero)
atomgroup.setFlags('pdbter', termini)
atomgroup.setAltlocs(altlocs)
atomgroup.setIcodes(np.char.strip(icodes))
atomgroup.setSerials(serials)
if isPDB:
if anisou is not None:
anisou.resize((acount, 6), refcheck=False)
atomgroup.setAnisous(anisou / 10000)
if siguij is not None:
siguij.resize((acount, 6), refcheck=False)
atomgroup.setAnistds(siguij / 10000)
bfactors.resize(acount, refcheck=False)
occupancies.resize(acount, refcheck=False)
segnames.resize(acount, refcheck=False)
elements.resize(acount, refcheck=False)
atomgroup.setSegnames(np.char.strip(segnames))
atomgroup.setElements(np.char.strip(elements))
from prody.utilities.misctools import getMasses
atomgroup.setMasses(getMasses(np.char.strip(elements)))
atomgroup.setBetas(bfactors)
atomgroup.setOccupancies(occupancies)
else:
charges.resize(acount, refcheck=False)
radii.resize(acount, refcheck=False)
atomgroup.setCharges(charges)
atomgroup.setRadii(radii)
if altloc and altloc_torf:
_evalAltlocs(atomgroup, altloc, chainids, resnums, resnames, atomnames)
return atomgroup
def getMasses(self):
"""get the mass atom. """
from prody.utilities.misctools import getMasses
return getMasses(self.getElement())
def _parsePDBLines(atomgroup,
lines,
split,
model,
chain,
subset,
altloc_torf,
format='PDB'):
"""Returns an AtomGroup. See also :func:`.parsePDBStream()`.
:arg lines: PDB/PQR lines
:arg split: starting index for coordinate data lines"""
format = format.upper()
if format == 'PDB':
isPDB = True
else:
isPDB = False
if subset:
if subset == 'ca':
subset = set(('CA', ))
elif subset in 'bb':
subset = flags.BACKBONE
only_subset = True
protein_resnames = flags.AMINOACIDS
else:
only_subset = False
if chain is None:
only_chains = False
else:
only_chains = True
onlycoords = False
n_atoms = atomgroup.numAtoms()
if n_atoms > 0:
asize = n_atoms
else:
# most PDB files contain less than 99999 atoms
asize = min(len(lines) - split, 99999)
addcoords = False
if atomgroup.numCoordsets() > 0:
addcoords = True
alength = asize
coordinates = np.zeros((asize, 3), dtype=float)
atomnames = np.zeros(asize, dtype=ATOMIC_FIELDS['name'].dtype)
resnames = np.zeros(asize, dtype=ATOMIC_FIELDS['resname'].dtype)
resnums = np.zeros(asize, dtype=ATOMIC_FIELDS['resnum'].dtype)
chainids = np.zeros(asize, dtype=ATOMIC_FIELDS['chain'].dtype)
hetero = np.zeros(asize, dtype=bool)
termini = np.zeros(asize, dtype=bool)
altlocs = np.zeros(asize, dtype=ATOMIC_FIELDS['altloc'].dtype)
icodes = np.zeros(asize, dtype=ATOMIC_FIELDS['icode'].dtype)
serials = np.zeros(asize, dtype=ATOMIC_FIELDS['serial'].dtype)
charges = np.zeros(asize, dtype=ATOMIC_FIELDS['charge'].dtype)
if isPDB:
segnames = np.zeros(asize, dtype=ATOMIC_FIELDS['segment'].dtype)
elements = np.zeros(asize, dtype=ATOMIC_FIELDS['element'].dtype)
bfactors = np.zeros(asize, dtype=ATOMIC_FIELDS['beta'].dtype)
occupancies = np.zeros(asize, dtype=ATOMIC_FIELDS['occupancy'].dtype)
anisou = None
siguij = None
else:
radii = np.zeros(asize, dtype=ATOMIC_FIELDS['radius'].dtype)
asize = 2000 # increase array length by this much when needed
start = split
stop = len(lines)
nmodel = 0
# if a specific model is requested, skip lines until that one
if isPDB and model is not None and model != 1:
for i in range(split, len(lines)):
if lines[i][:5] == 'MODEL':
nmodel += 1
if model == nmodel:
start = i + 1
stop = len(lines)
break
if nmodel != model:
raise PDBParseError('model {0} is not found'.format(model))
if isinstance(altloc_torf, str):
if altloc_torf.strip() != 'A':
LOGGER.info('Parsing alternate locations {0}.'.format(altloc_torf))
which_altlocs = ' ' + ''.join(altloc_torf.split())
else:
which_altlocs = ' A'
altloc_torf = False
else:
which_altlocs = ' A'
altloc_torf = True
acount = 0
coordsets = None
altloc = defaultdict(list)
i = start
END = False
while i < stop:
line = lines[i]
if not isPDB:
fields = line.split()
if len(fields) == 10:
fields.insert(4, '')
elif len(fields) != 11:
LOGGER.warn(
'wrong number of fields for PQR format at line %d' % i)
i += 1
continue
if isPDB:
startswith = line[0:6].strip()
else:
startswith = fields[0]
if startswith == 'ATOM' or startswith == 'HETATM':
if isPDB:
atomname = line[12:16].strip()
resname = line[17:21].strip()
else:
atomname = fields[2]
resname = fields[3]
if only_subset:
if not (atomname in subset and resname in protein_resnames):
i += 1
continue
if isPDB:
chid = line[21]
else:
chid = fields[4]
if only_chains:
if not chid in chain:
i += 1
continue
if isPDB:
alt = line[16]
if alt not in which_altlocs:
altloc[alt].append((line, i))
i += 1
continue
else:
alt = ' '
try:
if isPDB:
coordinates[acount, 0] = line[30:38]
coordinates[acount, 1] = line[38:46]
coordinates[acount, 2] = line[46:54]
else:
coordinates[acount, 0] = fields[6]
coordinates[acount, 1] = fields[7]
coordinates[acount, 2] = fields[8]
except:
if acount >= n_atoms > 0:
if nmodel == 0:
raise ValueError(format + 'file and AtomGroup ag must '
'have same number of atoms')
LOGGER.warn(
'Discarding model {0}, which contains {1} more '
'atoms than first model does.'.format(
nmodel + 1, acount - n_atoms + 1))
acount = 0
nmodel += 1
coordinates = np.zeros((n_atoms, 3), dtype=float)
if isPDB:
while lines[i][:6] != 'ENDMDL':
i += 1
else:
raise PDBParseError('invalid or missing coordinate(s) at '
'line {0}'.format(i + 1))
if onlycoords:
acount += 1
i += 1
continue
try:
serials[acount] = int(line[6:11]) if isPDB else int(fields[1])
except ValueError:
try:
serials[acount] = int(line[6:11], 16) if isPDB else int(
fields[1], 16)
except ValueError:
LOGGER.warn(
'failed to parse serial number in line {0}'.format(i))
serials[acount] = serials[acount - 1] + 1
altlocs[acount] = alt
atomnames[acount] = atomname
resnames[acount] = resname
chainids[acount] = chid
if isPDB:
resnums[acount] = line[22:26]
icodes[acount] = line[26]
else:
resnum = fields[5]
if resnum[-1].isalpha():
icode = resnum[-1]
else:
icode = ' '
resnums[acount] = resnum
icodes[acount] = icode
if isPDB:
try:
occupancies[acount] = line[54:60]
except:
LOGGER.warn(
'failed to parse occupancy at line {0}'.format(i))
try:
bfactors[acount] = line[60:66]
except:
LOGGER.warn(
'failed to parse beta-factor at line {0}'.format(i))
hetero[acount] = startswith[0] == 'H'
segnames[acount] = line[72:76]
elements[acount] = line[76:78]
try:
charges[acount] = int(line[79] + line[78])
except:
charges[acount] = 0
else:
try:
charges[acount] = fields[9]
except:
LOGGER.warn('failed to parse charge at line {0}'.format(i))
try:
radii[acount] = fields[10]
except:
LOGGER.warn('failed to parse radius at line {0}'.format(i))
acount += 1
if n_atoms == 0 and acount >= alength:
# if arrays are short extend them with zeros
alength += asize
coordinates = np.concatenate(
(coordinates, np.zeros((asize, 3), float)))
atomnames = np.concatenate(
(atomnames, np.zeros(asize, ATOMIC_FIELDS['name'].dtype)))
resnames = np.concatenate(
(resnames, np.zeros(asize,
ATOMIC_FIELDS['resname'].dtype)))
resnums = np.concatenate(
(resnums, np.zeros(asize, ATOMIC_FIELDS['resnum'].dtype)))
chainids = np.concatenate(
(chainids, np.zeros(asize, ATOMIC_FIELDS['chain'].dtype)))
hetero = np.concatenate((hetero, np.zeros(asize, bool)))
termini = np.concatenate((termini, np.zeros(asize, bool)))
altlocs = np.concatenate(
(altlocs, np.zeros(asize, ATOMIC_FIELDS['altloc'].dtype)))
icodes = np.concatenate(
(icodes, np.zeros(asize, ATOMIC_FIELDS['icode'].dtype)))
serials = np.concatenate(
(serials, np.zeros(asize, ATOMIC_FIELDS['serial'].dtype)))
if isPDB:
bfactors = np.concatenate(
(bfactors, np.zeros(asize,
ATOMIC_FIELDS['beta'].dtype)))
occupancies = np.concatenate(
(occupancies,
np.zeros(asize, ATOMIC_FIELDS['occupancy'].dtype)))
segnames = np.concatenate(
(segnames,
np.zeros(asize, ATOMIC_FIELDS['segment'].dtype)))
elements = np.concatenate(
(elements,
np.zeros(asize, ATOMIC_FIELDS['element'].dtype)))
if anisou is not None:
anisou = np.concatenate(
(anisou,
np.zeros((asize, 6),
ATOMIC_FIELDS['anisou'].dtype)))
if siguij is not None:
siguij = np.concatenate(
(siguij,
np.zeros((asize, 6),
ATOMIC_FIELDS['siguij'].dtype)))
else:
charges = np.concatenate(
(charges, np.zeros(asize,
ATOMIC_FIELDS['charge'].dtype)))
radii = np.concatenate(
(radii, np.zeros(asize,
ATOMIC_FIELDS['radius'].dtype)))
#elif startswith == 'END ' or startswith == 'CONECT':
# i += 1
# break
elif not onlycoords and (startswith == 'TER '
or startswith.strip() == 'TER'):
termini[acount - 1] = True
elif startswith == 'ENDMDL' or startswith[:3] == 'END':
if acount == 0:
# If there is no atom record between ENDMDL & END skip to next
i += 1
continue
if model is not None:
i += 1
break
diff = stop - i - 1
END = diff < acount
if coordsets is not None:
END = END or nmodel >= coordsets.shape[0]
if onlycoords:
if acount < n_atoms:
LOGGER.warn('Discarding model {0}, which contains '
'{1} fewer atoms than the first model '
'does.'.format(nmodel + 1, n_atoms - acount))
else:
coordsets[nmodel] = coordinates
nmodel += 1
acount = 0
if not END:
coordinates = coordsets[nmodel]
else:
if acount != n_atoms > 0:
raise ValueError('PDB file and AtomGroup ag must have '
'same number of atoms')
# this is where to decide if more coordsets should be expected
if END:
coordinates.resize((acount, 3), refcheck=False)
if addcoords:
atomgroup.addCoordset(coordinates)
else:
atomgroup._setCoords(coordinates)
else:
coordsets = np.zeros((int(diff // acount + 1), acount, 3))
coordsets[0] = coordinates[:acount]
onlycoords = True
atomnames.resize(acount, refcheck=False)
resnames.resize(acount, refcheck=False)
resnums.resize(acount, refcheck=False)
chainids.resize(acount, refcheck=False)
hetero.resize(acount, refcheck=False)
termini.resize(acount, refcheck=False)
altlocs.resize(acount, refcheck=False)
icodes.resize(acount, refcheck=False)
serials.resize(acount, refcheck=False)
if not only_subset:
atomnames = np.char.strip(atomnames)
resnames = np.char.strip(resnames)
atomgroup.setNames(atomnames)
atomgroup.setResnames(resnames)
atomgroup.setResnums(resnums)
atomgroup.setChids(chainids)
atomgroup.setFlags('hetatm', hetero)
atomgroup.setFlags('pdbter', termini)
atomgroup.setAltlocs(altlocs)
atomgroup.setIcodes(np.char.strip(icodes))
atomgroup.setSerials(serials)
if isPDB:
bfactors.resize(acount, refcheck=False)
occupancies.resize(acount, refcheck=False)
segnames.resize(acount, refcheck=False)
elements.resize(acount, refcheck=False)
atomgroup.setBetas(bfactors)
atomgroup.setOccupancies(occupancies)
atomgroup.setSegnames(np.char.strip(segnames))
atomgroup.setElements(np.char.strip(elements))
from prody.utilities.misctools import getMasses
atomgroup.setMasses(getMasses(np.char.strip(elements)))
if anisou is not None:
anisou.resize((acount, 6), refcheck=False)
atomgroup.setAnisous(anisou / 10000)
if siguij is not None:
siguij.resize((acount, 6), refcheck=False)
atomgroup.setAnistds(siguij / 10000)
else:
charges.resize(acount, refcheck=False)
radii.resize(acount, refcheck=False)
atomgroup.setCharges(charges)
atomgroup.setRadii(radii)
nmodel += 1
n_atoms = acount
acount = 0
coordinates = np.zeros((n_atoms, 3), dtype=float)
if altloc and altloc_torf:
_evalAltlocs(atomgroup, altloc, chainids, resnums,
resnames, atomnames)
altloc = defaultdict(list)
if END:
break
elif isPDB and startswith == 'ANISOU':
if anisou is None:
anisou = True
anisou = np.zeros((alength, 6),
dtype=ATOMIC_FIELDS['anisou'].dtype)
try:
index = acount - 1
anisou[index, 0] = line[28:35]
anisou[index, 1] = line[35:42]
anisou[index, 2] = line[43:49]
anisou[index, 3] = line[49:56]
anisou[index, 4] = line[56:63]
anisou[index, 5] = line[63:70]
except:
LOGGER.warn('failed to parse anisotropic temperature '
'factors at line {0}'.format(i))
elif isPDB and startswith == 'SIGUIJ':
if siguij is None:
siguij = np.zeros((alength, 6),
dtype=ATOMIC_FIELDS['siguij'].dtype)
try:
index = acount - 1
siguij[index, 0] = line[28:35]
siguij[index, 1] = line[35:42]
siguij[index, 2] = line[43:49]
siguij[index, 3] = line[49:56]
siguij[index, 4] = line[56:63]
siguij[index, 5] = line[63:70]
except:
LOGGER.warn(
'failed to parse standard deviations of '
'anisotropic temperature factors at line {0}'.format(i))
elif startswith == 'SIGATM':
pass
i += 1
if onlycoords:
if acount == atomgroup.numAtoms():
coordsets[nmodel] = coordinates
nmodel += 1
del coordinates
coordsets.resize((nmodel, atomgroup.numAtoms(), 3), refcheck=False)
if addcoords:
atomgroup.addCoordset(coordsets)
else:
atomgroup._setCoords(coordsets)
elif not END:
# this means last line was an ATOM line, so atomgroup is not decorated
coordinates.resize((acount, 3), refcheck=False)
if addcoords:
atomgroup.addCoordset(coordinates)
else:
atomgroup._setCoords(coordinates)
atomnames.resize(acount, refcheck=False)
resnames.resize(acount, refcheck=False)
resnums.resize(acount, refcheck=False)
chainids.resize(acount, refcheck=False)
hetero.resize(acount, refcheck=False)
termini.resize(acount, refcheck=False)
altlocs.resize(acount, refcheck=False)
icodes.resize(acount, refcheck=False)
serials.resize(acount, refcheck=False)
if not only_subset:
atomnames = np.char.strip(atomnames)
resnames = np.char.strip(resnames)
atomgroup.setNames(atomnames)
atomgroup.setResnames(resnames)
atomgroup.setResnums(resnums)
atomgroup.setChids(chainids)
atomgroup.setFlags('hetatm', hetero)
atomgroup.setFlags('pdbter', termini)
atomgroup.setAltlocs(altlocs)
atomgroup.setIcodes(np.char.strip(icodes))
atomgroup.setSerials(serials)
if isPDB:
if anisou is not None:
anisou.resize((acount, 6), refcheck=False)
atomgroup.setAnisous(anisou / 10000)
if siguij is not None:
siguij.resize((acount, 6), refcheck=False)
atomgroup.setAnistds(siguij / 10000)
bfactors.resize(acount, refcheck=False)
occupancies.resize(acount, refcheck=False)
segnames.resize(acount, refcheck=False)
elements.resize(acount, refcheck=False)
atomgroup.setSegnames(np.char.strip(segnames))
atomgroup.setElements(np.char.strip(elements))
from prody.utilities.misctools import getMasses
atomgroup.setMasses(getMasses(np.char.strip(elements)))
atomgroup.setBetas(bfactors)
atomgroup.setOccupancies(occupancies)
else:
charges.resize(acount, refcheck=False)
radii.resize(acount, refcheck=False)
atomgroup.setCharges(charges)
atomgroup.setRadii(radii)
if altloc and altloc_torf:
_evalAltlocs(atomgroup, altloc, chainids, resnums, resnames, atomnames)
return atomgroup
def _parseMMCIFLines(atomgroup, lines, model, chain, subset, altloc_torf,
header):
"""Returns an AtomGroup. See also :func:`.parsePDBStream()`.
:arg lines: mmCIF lines
"""
if subset is not None:
if subset == 'ca':
subset = set(('CA', ))
elif subset in 'bb':
subset = flags.BACKBONE
protein_resnames = flags.AMINOACIDS
asize = 0
i = 0
models = []
nModels = 0
fields = {}
fieldCounter = -1
foundAtomBlock = False
doneAtomBlock = False
while not doneAtomBlock:
line = lines[i]
if line[:11] == '_atom_site.':
fieldCounter += 1
fields[line.split('.')[1].strip()] = fieldCounter
if line.startswith('ATOM') or line.startswith('HETATM'):
if not foundAtomBlock:
foundAtomBlock = True
start = i
models.append(line.split()[fields['pdbx_PDB_model_num']])
if models[asize] != models[asize - 1]:
nModels += 1
asize += 1
else:
if foundAtomBlock:
doneAtomBlock = True
stop = i
i += 1
if nModels == 0:
nModels = 1
if model is not None and model != 1:
for i in range(start, stop):
if str(models[i]) != model and str(models[i + 1]) == model:
start = i + 1
if str(models[i]) == model and str(models[i + 1]) != model:
stop = i + 1
break
if not str(model) in models:
raise mmCIFParseError('model {0} is not found'.format(model))
addcoords = False
if atomgroup.numCoordsets() > 0:
addcoords = True
if isinstance(altloc_torf, str):
if altloc_torf.strip() != 'A':
LOGGER.info('Parsing alternate locations {0}.'.format(altloc_torf))
which_altlocs = '.' + ''.join(altloc_torf.split())
else:
which_altlocs = '.A'
altloc_torf = False
else:
which_altlocs = '.A'
altloc_torf = True
coordinates = np.zeros((asize, 3), dtype=float)
atomnames = np.zeros(asize, dtype=ATOMIC_FIELDS['name'].dtype)
resnames = np.zeros(asize, dtype=ATOMIC_FIELDS['resname'].dtype)
resnums = np.zeros(asize, dtype=ATOMIC_FIELDS['resnum'].dtype)
chainids = np.zeros(asize, dtype=ATOMIC_FIELDS['chain'].dtype)
segnames = np.zeros(asize, dtype=ATOMIC_FIELDS['segment'].dtype)
hetero = np.zeros(asize, dtype=bool)
termini = np.zeros(asize, dtype=bool)
altlocs = np.zeros(asize, dtype=ATOMIC_FIELDS['altloc'].dtype)
icodes = np.zeros(asize, dtype=ATOMIC_FIELDS['icode'].dtype)
serials = np.zeros(asize, dtype=ATOMIC_FIELDS['serial'].dtype)
elements = np.zeros(asize, dtype=ATOMIC_FIELDS['element'].dtype)
bfactors = np.zeros(asize, dtype=ATOMIC_FIELDS['beta'].dtype)
occupancies = np.zeros(asize, dtype=ATOMIC_FIELDS['occupancy'].dtype)
n_atoms = atomgroup.numAtoms()
if n_atoms > 0:
asize = n_atoms
acount = 0
for line in lines[start:stop]:
startswith = line.split()[fields['group_PDB']]
atomname = line.split()[fields['auth_atom_id']]
resname = line.split()[fields['auth_comp_id']]
if subset is not None:
if not (atomname in subset and resname in protein_resnames):
continue
chID = line.split()[fields['auth_asym_id']]
if chain is not None:
if isinstance(chain, str):
chain = chain.split(',')
if not chID in chain:
continue
segID = line.split()[fields['label_asym_id']]
alt = line.split()[fields['label_alt_id']]
if alt not in which_altlocs:
continue
if model is not None:
if int(models[acount]) < model:
continue
elif int(models[acount]) > model:
break
coordinates[acount] = [
line.split()[fields['Cartn_x']],
line.split()[fields['Cartn_y']],
line.split()[fields['Cartn_z']]
]
atomnames[acount] = atomname
resnames[acount] = resname
resnums[acount] = line.split()[fields['auth_seq_id']]
chainids[acount] = chID
segnames[acount] = segID
hetero[acount] = startswith == 'HETATM' # True or False
if chainids[acount] != chainids[acount - 1]:
termini[acount - 1] = True
altlocs[acount] = alt
icodes[acount] = line.split()[fields['pdbx_PDB_ins_code']]
if icodes[acount] == '?':
icodes[acount] = ''
serials[acount] = line.split()[fields['id']]
elements[acount] = line.split()[fields['type_symbol']]
bfactors[acount] = line.split()[fields['B_iso_or_equiv']]
occupancies[acount] = line.split()[fields['occupancy']]
acount += 1
if model is not None:
nModels = 1
modelSize = acount // nModels
if addcoords:
atomgroup.addCoordset(coordinates[:modelSize])
else:
atomgroup._setCoords(coordinates[:modelSize])
atomgroup.setNames(atomnames[:modelSize])
atomgroup.setResnames(resnames[:modelSize])
atomgroup.setResnums(resnums[:modelSize])
atomgroup.setSegnames(segnames[:modelSize])
atomgroup.setChids(chainids[:modelSize])
atomgroup.setFlags('hetatm', hetero[:modelSize])
atomgroup.setFlags('pdbter', termini[:modelSize])
atomgroup.setAltlocs(altlocs[:modelSize])
atomgroup.setIcodes(icodes[:modelSize])
atomgroup.setSerials(serials[:modelSize])
atomgroup.setElements(elements[:modelSize])
from prody.utilities.misctools import getMasses
atomgroup.setMasses(getMasses(elements[:modelSize]))
atomgroup.setBetas(bfactors[:modelSize])
atomgroup.setOccupancies(occupancies[:modelSize])
for n in range(1, nModels):
atomgroup.addCoordset(coordinates[n * modelSize:(n + 1) * modelSize])
if header:
header = parseSTARLines(lines[:start - fieldCounter - 2] +
lines[stop:],
shlex=True)
return atomgroup, header
return atomgroup
def _parseMMCIFLines(atomgroup, lines, model, chain, subset,
altloc_torf):
"""Returns an AtomGroup. See also :func:`.parsePDBStream()`.
:arg lines: mmCIF lines
"""
if subset is not None:
if subset == 'ca':
subset = set(('CA',))
elif subset in 'bb':
subset = flags.BACKBONE
protein_resnames = flags.AMINOACIDS
asize = 0
i = 0
models = []
nModels = 0
fields = OrderedDict()
fieldCounter = -1
foundAtomBlock = False
doneAtomBlock = False
start = 0
stop = 0
while not doneAtomBlock:
line = lines[i]
if line[:11] == '_atom_site.':
fieldCounter += 1
fields[line.split('.')[1].strip()] = fieldCounter
if line.startswith('ATOM ') or line.startswith('HETATM'):
if not foundAtomBlock:
foundAtomBlock = True
start = i
models.append(line.split()[fields['pdbx_PDB_model_num']])
if len(models) == 1 or (models[asize] != models[asize-1]):
nModels += 1
asize += 1
else:
if foundAtomBlock:
doneAtomBlock = True
stop = i
i += 1
if model is not None and model != 1:
for i in range(start, stop):
if str(models[i]) != model and str(models[i+1]) == model:
start = i+1
if str(models[i]) == model and str(models[i+1]) != model:
stop = i+1
break
if not str(model) in models:
raise mmCIFParseError('model {0} is not found'.format(model))
addcoords = False
if atomgroup.numCoordsets() > 0:
addcoords = True
if isinstance(altloc_torf, str):
if altloc_torf == 'all':
which_altlocs = 'all'
elif altloc_torf.strip() != 'A':
LOGGER.info('Parsing alternate locations {0}.'
.format(altloc_torf))
which_altlocs = '.' + ''.join(altloc_torf.split())
else:
which_altlocs = '.A'
altloc_torf = False
else:
which_altlocs = '.A'
altloc_torf = True
coordinates = np.zeros((asize, 3), dtype=float)
atomnames = np.zeros(asize, dtype=ATOMIC_FIELDS['name'].dtype)
resnames = np.zeros(asize, dtype=ATOMIC_FIELDS['resname'].dtype)
resnums = np.zeros(asize, dtype=ATOMIC_FIELDS['resnum'].dtype)
chainids = np.zeros(asize, dtype=ATOMIC_FIELDS['chain'].dtype)
segnames = np.zeros(asize, dtype=ATOMIC_FIELDS['segment'].dtype)
hetero = np.zeros(asize, dtype=bool)
termini = np.zeros(asize, dtype=bool)
altlocs = np.zeros(asize, dtype=ATOMIC_FIELDS['altloc'].dtype)
icodes = np.zeros(asize, dtype=ATOMIC_FIELDS['icode'].dtype)
serials = np.zeros(asize, dtype=ATOMIC_FIELDS['serial'].dtype)
elements = np.zeros(asize, dtype=ATOMIC_FIELDS['element'].dtype)
bfactors = np.zeros(asize, dtype=ATOMIC_FIELDS['beta'].dtype)
occupancies = np.zeros(asize, dtype=ATOMIC_FIELDS['occupancy'].dtype)
n_atoms = atomgroup.numAtoms()
if n_atoms > 0:
asize = n_atoms
acount = 0
for line in lines[start:stop]:
startswith = line.split()[fields['group_PDB']]
atomname = line.split()[fields['auth_atom_id']]
if atomname.startswith('"') and atomname.endswith('"'):
atomname = atomname[1:-1]
resname = line.split()[fields['auth_comp_id']]
if subset is not None:
if not (atomname in subset and resname in protein_resnames):
continue
chID = line.split()[fields['auth_asym_id']]
if chain is not None:
if isinstance(chain, str):
chain = chain.split(',')
if not chID in chain:
continue
segID = line.split()[fields['label_asym_id']]
alt = line.split()[fields['label_alt_id']]
if alt not in which_altlocs and which_altlocs != 'all':
continue
if alt == '.':
alt = ' '
if model is not None:
if int(models[acount]) < model:
continue
elif int(models[acount]) > model:
break
coordinates[acount] = [line.split()[fields['Cartn_x']],
line.split()[fields['Cartn_y']],
line.split()[fields['Cartn_z']]]
atomnames[acount] = atomname
resnames[acount] = resname
resnums[acount] = line.split()[fields['auth_seq_id']]
chainids[acount] = chID
segnames[acount] = segID
hetero[acount] = startswith == 'HETATM' # True or False
if chainids[acount] != chainids[acount-1]:
termini[acount-1] = True
altlocs[acount] = alt
icodes[acount] = line.split()[fields['pdbx_PDB_ins_code']]
if icodes[acount] == '?':
icodes[acount] = ''
serials[acount] = line.split()[fields['id']]
elements[acount] = line.split()[fields['type_symbol']]
bfactors[acount] = line.split()[fields['B_iso_or_equiv']]
occupancies[acount] = line.split()[fields['occupancy']]
acount += 1
if model is not None:
nModels = 1
modelSize = acount//nModels
if addcoords:
atomgroup.addCoordset(coordinates[:modelSize])
else:
atomgroup._setCoords(coordinates[:modelSize])
atomgroup.setNames(atomnames[:modelSize])
atomgroup.setResnames(resnames[:modelSize])
atomgroup.setResnums(resnums[:modelSize])
atomgroup.setSegnames(segnames[:modelSize])
atomgroup.setChids(chainids[:modelSize])
atomgroup.setFlags('hetatm', hetero[:modelSize])
atomgroup.setFlags('pdbter', termini[:modelSize])
atomgroup.setAltlocs(altlocs[:modelSize])
atomgroup.setIcodes(icodes[:modelSize])
atomgroup.setSerials(serials[:modelSize])
atomgroup.setElements(elements[:modelSize])
from prody.utilities.misctools import getMasses
atomgroup.setMasses(getMasses(elements[:modelSize]))
atomgroup.setBetas(bfactors[:modelSize])
atomgroup.setOccupancies(occupancies[:modelSize])
anisou = None
siguij = None
try:
data = parseSTARSection(lines, "_atom_site_anisotrop")
x = data[0] # check if data has anything in it
except IndexError:
LOGGER.warn("No anisotropic B factors found")
else:
anisou = np.zeros((acount, 6),
dtype=ATOMIC_FIELDS['anisou'].dtype)
if "_atom_site_anisotrop.U[1][1]_esd" in data[0].keys():
siguij = np.zeros((alength, 6),
dtype=ATOMIC_FIELDS['siguij'].dtype)
for entry in data:
try:
index = np.where(atomgroup.getSerials() == int(
entry["_atom_site_anisotrop.id"]))[0][0]
except:
continue
anisou[index, 0] = entry['_atom_site_anisotrop.U[1][1]']
anisou[index, 1] = entry['_atom_site_anisotrop.U[2][2]']
anisou[index, 2] = entry['_atom_site_anisotrop.U[3][3]']
anisou[index, 3] = entry['_atom_site_anisotrop.U[1][2]']
anisou[index, 4] = entry['_atom_site_anisotrop.U[1][3]']
anisou[index, 5] = entry['_atom_site_anisotrop.U[2][3]']
if siguij is not None:
siguij[index, 0] = entry['_atom_site_anisotrop.U[1][1]_esd']
siguij[index, 1] = entry['_atom_site_anisotrop.U[2][2]_esd']
siguij[index, 2] = entry['_atom_site_anisotrop.U[3][3]_esd']
siguij[index, 3] = entry['_atom_site_anisotrop.U[1][2]_esd']
siguij[index, 4] = entry['_atom_site_anisotrop.U[1][3]_esd']
siguij[index, 5] = entry['_atom_site_anisotrop.U[2][3]_esd']
atomgroup.setAnisous(anisou) # no division needed anymore
atomgroup.setAnistds(siguij) # no division needed anymore
for n in range(1, nModels):
atomgroup.addCoordset(coordinates[n*modelSize:(n+1)*modelSize])
return atomgroup
def getMasses(self):
"""get the mass atom. """
from prody.utilities.misctools import getMasses
return getMasses(self.getElement())
def _parseCIFLines(atomgroup, lines, model, chain, subset,
altloc_torf):
"""Returns an AtomGroup. See also :func:`.parsePDBStream()`.
:arg lines: CIF lines
"""
if subset is not None:
if subset == 'ca':
subset = set(('CA',))
elif subset in 'bb':
subset = flags.BACKBONE
protein_resnames = flags.AMINOACIDS
asize = 0
i = 0
models = []
nModels = 0
fields = {}
fieldCounter = -1
foundModelNumFieldID = False
foundAtomBlock = False
doneAtomBlock = False
while not doneAtomBlock:
line = lines[i]
if line[:11] == '_atom_site.':
fieldCounter += 1
fields[line.split('.')[1].strip()] = fieldCounter
if line.startswith('ATOM') or line.startswith('HETATM'):
if not foundAtomBlock:
foundAtomBlock = True
start = i
models.append(line.split()[fields['pdbx_PDB_model_num']])
if models[asize] != models[asize-1]:
nModels += 1
asize += 1
else:
if foundAtomBlock:
doneAtomBlock = True
i += 1
stop = i-1
if nModels == 0: nModels = 1
if model is not None and model != 1:
for i in range(start, stop):
if str(models[i]) != model and str(models[i+1]) == model:
start = i+1
if str(models[i]) == model and str(models[i+1]) != model:
stop = i+1
break
if not str(model) in models:
raise CIFParseError('model {0} is not found'.format(model))
addcoords = False
if atomgroup.numCoordsets() > 0:
addcoords = True
if isinstance(altloc_torf, str):
if altloc_torf.strip() != 'A':
LOGGER.info('Parsing alternate locations {0}.'
.format(altloc_torf))
which_altlocs = '.' + ''.join(altloc_torf.split())
else:
which_altlocs = '.A'
altloc_torf = False
else:
which_altlocs = '.A'
altloc_torf = True
coordinates = np.zeros((asize, 3), dtype=float)
atomnames = np.zeros(asize, dtype=ATOMIC_FIELDS['name'].dtype)
resnames = np.zeros(asize, dtype=ATOMIC_FIELDS['resname'].dtype)
resnums = np.zeros(asize, dtype=ATOMIC_FIELDS['resnum'].dtype)
chainids = np.zeros(asize, dtype=ATOMIC_FIELDS['chain'].dtype)
hetero = np.zeros(asize, dtype=bool)
termini = np.zeros(asize, dtype=bool)
altlocs = np.zeros(asize, dtype=ATOMIC_FIELDS['altloc'].dtype)
icodes = np.zeros(asize, dtype=ATOMIC_FIELDS['icode'].dtype)
serials = np.zeros(asize, dtype=ATOMIC_FIELDS['serial'].dtype)
elements = np.zeros(asize, dtype=ATOMIC_FIELDS['element'].dtype)
bfactors = np.zeros(asize, dtype=ATOMIC_FIELDS['beta'].dtype)
occupancies = np.zeros(asize, dtype=ATOMIC_FIELDS['occupancy'].dtype)
n_atoms = atomgroup.numAtoms()
if n_atoms > 0:
asize = n_atoms
acount = 0
for line in lines[start:stop]:
startswith = line.split()[fields['group_PDB']]
atomname = line.split()[fields['auth_atom_id']]
resname = line.split()[fields['auth_comp_id']]
if subset is not None:
if not (atomname in subset and resname in protein_resnames):
continue
chID = line.split()[fields['auth_asym_id']]
if chain is not None:
if not chID in chain:
continue
alt = line.split()[fields['label_alt_id']]
if alt not in which_altlocs:
continue
if model is not None:
if int(models[acount]) < model:
continue
elif int(models[acount]) > model:
break
coordinates[acount] = [line.split()[fields['Cartn_x']], \
line.split()[fields['Cartn_y']], \
line.split()[fields['Cartn_z']]]
atomnames[acount] = atomname
resnames[acount] = resname
resnums[acount] = line.split()[fields['auth_seq_id']]
chainids[acount] = chID
hetero[acount] = startswith == 'HETATM' # True or False
if chainids[acount] != chainids[acount-1]: termini[acount] = True
altlocs[acount] = alt
icodes[acount] = line.split()[fields['pdbx_PDB_ins_code']]
if icodes[acount] == '?': icodes[acount] = ''
serials[acount] = line.split()[fields['id']]
elements[acount] = line.split()[fields['type_symbol']]
bfactors[acount] = line.split()[fields['B_iso_or_equiv']]
occupancies[acount] = line.split()[fields['occupancy']]
acount += 1
if model is not None:
nModels = 1
modelSize = acount//nModels
if addcoords:
atomgroup.addCoordset(coordinates[:modelSize])
else:
atomgroup._setCoords(coordinates[:modelSize])
atomgroup.setNames(atomnames[:modelSize])
atomgroup.setResnames(resnames[:modelSize])
atomgroup.setResnums(resnums[:modelSize])
atomgroup.setChids(chainids[:modelSize])
atomgroup.setFlags('hetatm', hetero[:modelSize])
atomgroup.setFlags('pdbter', termini[:modelSize])
atomgroup.setAltlocs(altlocs[:modelSize])
atomgroup.setIcodes(icodes[:modelSize])
atomgroup.setSerials(serials[:modelSize])
atomgroup.setElements(elements[:modelSize])
from prody.utilities.misctools import getMasses
atomgroup.setMasses(getMasses(elements[:modelSize]))
atomgroup.setBetas(bfactors[:modelSize])
atomgroup.setOccupancies(occupancies[:modelSize])
for n in range(1,nModels):
atomgroup.addCoordset(coordinates[n*modelSize:(n+1)*modelSize])
return atomgroup
def _parseCIFLines(atomgroup, lines, model, chain, subset,
altloc_torf):
"""Returns an AtomGroup. See also :func:`.parsePDBStream()`.
:arg lines: CIF lines
"""
if subset is not None:
if subset == 'ca':
subset = set(('CA',))
elif subset in 'bb':
subset = flags.BACKBONE
protein_resnames = flags.AMINOACIDS
asize = 0
i = 0
models = []
nModels = 0
foundAtomBlock = False
doneAtomBlock = False
while not doneAtomBlock:
line = lines[i]
if line[:6] == 'ATOM ' or line[:6] == 'HETATM':
if not foundAtomBlock:
foundAtomBlock = True
start = i
models.append(line.split()[25]) # pdbx_PDB_model_num
if models[asize] != models[asize-1]:
nModels += 1
asize += 1
else:
if foundAtomBlock:
doneAtomBlock = True
i += 1
stop = i-1
if nModels == 0: nModels = 1
if model is not None and model != 1:
for i in range(start, stop):
if str(models[i]) != model and str(models[i+1]) == model:
start = i+1
if str(models[i]) == model and str(models[i+1]) != model:
stop = i+1
break
if not str(model) in models:
raise CIFParseError('model {0} is not found'.format(model))
addcoords = False
if atomgroup.numCoordsets() > 0:
addcoords = True
if isinstance(altloc_torf, str):
if altloc_torf.strip() != 'A':
LOGGER.info('Parsing alternate locations {0}.'
.format(altloc_torf))
which_altlocs = '.' + ''.join(altloc_torf.split())
else:
which_altlocs = '.A'
altloc_torf = False
else:
which_altlocs = '.A'
altloc_torf = True
coordinates = np.zeros((asize, 3), dtype=float)
atomnames = np.zeros(asize, dtype=ATOMIC_FIELDS['name'].dtype)
resnames = np.zeros(asize, dtype=ATOMIC_FIELDS['resname'].dtype)
resnums = np.zeros(asize, dtype=ATOMIC_FIELDS['resnum'].dtype)
chainids = np.zeros(asize, dtype=ATOMIC_FIELDS['chain'].dtype)
hetero = np.zeros(asize, dtype=bool)
termini = np.zeros(asize, dtype=bool)
altlocs = np.zeros(asize, dtype=ATOMIC_FIELDS['altloc'].dtype)
icodes = np.zeros(asize, dtype=ATOMIC_FIELDS['icode'].dtype)
serials = np.zeros(asize, dtype=ATOMIC_FIELDS['serial'].dtype)
elements = np.zeros(asize, dtype=ATOMIC_FIELDS['element'].dtype)
bfactors = np.zeros(asize, dtype=ATOMIC_FIELDS['beta'].dtype)
occupancies = np.zeros(asize, dtype=ATOMIC_FIELDS['occupancy'].dtype)
n_atoms = atomgroup.numAtoms()
if n_atoms > 0:
asize = n_atoms
acount = 0
for line in lines[start:stop]:
startswith = line.split()[0] # group_PDB
atomname = line.split()[-2] # auth_atom_id in stardard pos
resname = line.split()[-4] # auth_comp_id in standard pos
if subset is not None:
if not (atomname in subset and resname in protein_resnames):
continue
chID = line.split()[-3] # auth_asym_id in stardard pos
if chain is not None:
if not chID in chain:
LOGGER.info('The loop has entered the chID continue block!!')
continue
alt = line.split()[4] # label_alt_id in standard pos
if alt not in which_altlocs:
LOGGER.info('The loop has entered the alt continue block!!')
LOGGER.info('line = {0}'.format(line))
continue
if model is not None:
if int(models[acount]) < model:
LOGGER.info('The loop has entered the model continue block!!')
continue
elif int(models[acount]) > model:
LOGGER.info('The loop has entered the model break block!!')
break
coordinates[acount] = line.split()[10:13]
atomnames[acount] = atomname
resnames[acount] = resname
resnums[acount] = line.split()[21] # auth_seq_id
chainids[acount] = chID
hetero[acount] = startswith == 'HETATM' # True or False
if chainids[acount] != chainids[acount-1]: termini[acount] = True
altlocs[acount] = alt
icodes[acount] = line.split()[9] # pdbx_PDB_ins_code
if icodes[acount] == '?': icodes[acount] = ''
serials[acount] = line.split()[1] # id
elements[acount] = line.split()[2] # type_symbol
bfactors[acount] = line.split()[14]
occupancies[acount] = line.split()[13]
acount += 1
if model is not None:
nModels = 1
modelSize = acount//nModels
if addcoords:
atomgroup.addCoordset(coordinates[:modelSize])
else:
atomgroup._setCoords(coordinates[:modelSize])
atomgroup.setNames(atomnames[:modelSize])
atomgroup.setResnames(resnames[:modelSize])
atomgroup.setResnums(resnums[:modelSize])
atomgroup.setChids(chainids[:modelSize])
atomgroup.setFlags('hetatm', hetero[:modelSize])
atomgroup.setFlags('pdbter', termini[:modelSize])
atomgroup.setAltlocs(altlocs[:modelSize])
atomgroup.setIcodes(icodes[:modelSize])
atomgroup.setSerials(serials[:modelSize])
atomgroup.setElements(elements[:modelSize])
from prody.utilities.misctools import getMasses
atomgroup.setMasses(getMasses(elements[:modelSize]))
atomgroup.setBetas(bfactors[:modelSize])
atomgroup.setOccupancies(occupancies[:modelSize])
for n in range(1,nModels):
atomgroup.addCoordset(coordinates[n*modelSize:(n+1)*modelSize])
return atomgroup
