def __init__(self, pdbFilename=None, **kwargs):
super(NatQ, self).__init__(pdbFilename, **kwargs)
# kwargs
self.nativeRad = kwargs.get('nativerad', 8.)
# Main
if pdbFilename is not None:
self.aa = PDBFile(self.pdbFilename)[0]
self.aa.parse()
self.cg = KTGo(self.aa)
self.nativeContacts = self.cg.get_nativeSCSC()
def parse(pdbFilename, **kwargs):
"""
Parse a *.pdb* plain text file into its constituent chains and segments, and
print one CHARMM formatted *.pdb* file per chain/segment combination.
*kwarg defaults are listed first*
**kwargs:**
| ``informat`` ['auto', 'pdborg', 'charmm'] # 'auto' -> detects input formatting
| ``outformat`` ['charmm', 'pdborg', 'auto'] # 'auto' -> same as input formatting
| ``outpath`` ['auto', user_specified_path] # 'auto' -> same as pdbFilename path
| ``fix_chainid`` [True, False]
| ``autofix`` [True, False]
| ``modelnum`` ['auto', 0, 1, 2, ...] # 'auto' -> uses the first model found
| ``pickleme`` [False, True] # pickles the PDBFile object for later use
| ``verbose`` [False, True]
| ``old_resid`` [False, True]
>>> parse('~/pychm/1yjp/1yjp.pdb',outpath='~',pickleme=True)
"""
# kwargs
kwargs = lowerKeys(kwargs)
inFormat = kwargs.get('informat', 'auto')
outFormat = kwargs.get('outformat', 'charmm')
outPath = kwargs.get('outpath', 'auto')
fix_chainid = kwargs.get('fix_chainid', True)
autoFix = kwargs.get('autofix', True)
modelNum = kwargs.get('modelnum', 'auto')
pickleMe = kwargs.get('pickleme', False)
verbose = kwargs.get('verbose', False)
old_resid = kwargs.get('old_resid', False)
# Repackage the PDBFile kwargs, make pdbFile object
pdbFileArgs = {'informat':inFormat, 'fix_chainid':fix_chainid,
'autofix':autoFix, 'verbose':verbose}
pdb = PDBFile(pdbFilename, **pdbFileArgs)
if verbose:
print '%s: Output formatting set to `%s`' % (pdb.code, outFormat)
# Get model number...
if modelNum == 'auto':
thisMol = pdb.iter_models().next()
else:
thisMol = pdb[modelNum]
if verbose:
print '%s: Loading `%s`' % (pdb.code, thisMol.name)
# Determine explicit output Path
if outPath == 'auto':
outPath = pdb.path
else:
outPath = expandPath(outPath)
mkdir(outPath)
if verbose:
print '%s: Output path set to `%s`' % (pdb.code, outPath)
# Do Work
thisMol.parse()
if verbose:
print '%s: Parsing `%s`' % (pdb.code, thisMol.name)
# Write CHARMMing style output.
segDict = {'nuc':'nuc', 'pro':'pro', 'good':'goodhet', 'bad':'het',
'dna':'dna', 'rna':'rna'}
stdoutList = []
# Write pdb files
writeArgs = {'outformat':outFormat, 'ter':True, 'end':False,
'old_resid':old_resid}
for seg in thisMol.iter_seg():
stdoutList.append('%s-%s' % (seg.chainid, segDict[seg.segType]))
name = '%s/new_%s-%s-%s.pdb' % (outPath, thisMol.code, seg.chainid,
segDict[seg.segType])
if verbose:
print '%s: Writing output to file `%s`' % (pdb.code, name)
seg.write(filename=name, **writeArgs)
# Write pickle (chk) file
if pickleMe:
pickleFilename = '%s/%s.chk' % (outPath, pdb.code)
pickleFile = open(pickleFilename,'w')
dump(pdb,pickleFile)
pickleFile.close()
if verbose:
print '%s: Writing pickle to file `%s`' % (pdb.code, pickleFilename)
if verbose:
print '\n\nEnd of verbosity\n\n'
# To STDOUT
print 'natom=%d' % len(thisMol)
print 'nwarn=%d' % len(thisMol.warnings)
if thisMol.warnings:
print 'warnings=%r' % thisMol.warnings
print 'seg=%r' % stdoutList
class NatQ(BaseAnalysis):
"""
DOCME
"""
def __init__(self, pdbFilename=None, **kwargs):
super(NatQ, self).__init__(pdbFilename, **kwargs)
# kwargs
self.nativeRad = kwargs.get('nativerad', 8.)
# Main
if pdbFilename is not None:
self.aa = PDBFile(self.pdbFilename)[0]
self.aa.parse()
self.cg = KTGo(self.aa)
self.nativeContacts = self.cg.get_nativeSCSC()
@Property
def data():
doc =\
"""
"""
def fget(self):
filename = '%s/natq.pickle' % self.anlPathname
try:
self._data = pickle.load(open(filename))
print 'found pickled data in: %s' % filename
return self._data
except IOError:
print 'processing data in: %s' % self.anlPathname
self._data = self.build_nativeContactMatrix()
pickle.dump(self._data, open(filename, 'w'))
return self._data
def fset(self, value):
assert isinstance(value, np.ndarray)
self._data = value
def fdel(self):
filename = '%s/natq.pickle' % self.anlPathname
del self._data
try:
os.remove(filename)
except IOError:
pass
return locals()
@Property
def inpPathname():
doc =\
"""
"""
def fget(self):
return self._inpPathname
def fset(self, value):
self._inpPathname = self.expandPath(value)
return locals()
@Property
def nativeRad():
doc =\
"""
The cutoff distance between two CG sidechain atom objects which is used
to determine if their parent residues are in 'contact'.
"""
def fget(self):
return self._nativeRad
def fset(self, value):
if value <= 1:
raise ValueError("The native contact radius must be at least 1.")
self._nativeRad = float(value)
return locals()
def build_nativeContactMatrix(self):
"""
Builds a 2D numpy array from available correl output files. The first
index references the contact number, the second index references the
correl time array value (which in turn corresponds to nstep of the
original dynamics in quanta of ``correlSkip``. This method can take
awhile because of disk IO.
"""
print "Building native contact matrix."
rowLength = (self.correlStop - self.correlStart) / self.correlSkip + 1
tmp = []
for i in xrange(len(self.nativeContacts)):
try:
tmp.append(load_correlOutput('%s/natq%04d.anl' % (self.anlPathname, i)))
except IOError:
tmp.append(np.zeros(rowLength))
print "Can't find correl output for natq number: %04d." % i
return np.array(tmp, dtype=np.float64)
def get_natQofT(self):
"""
"""
data = self.data
tmp = np.zeros(data.shape)
tmp[data <= self.nativeRad] = 1
tmp[data > self.nativeRad] = 0
return tmp.mean(axis=0)
# Charmm input creation
def do_correl(self, **kwargs):
self.write_correlInput(**kwargs)
self.run_correlInput()
def write_correlInput(self, **kwargs):
kwargs = lowerKeys(kwargs)
header = kwargs.get('header', [])
#
print self.anlPathname
if self.anlPathname is None:
raise IOError("No directory specified.")
mkdir(self.anlPathname)
#
for i, group in enumerate(grouper(self.nativeContacts, 100)):
String = self.get_correlInputHeader(header)
String.append('! anl :: write')
for j, contact in enumerate(group):
String.append('open unit %03d write card name %s/natq%02d%02d.anl' % (j+100, self.anlPathname, i, j))
String.append('')
String.append('correl maxtimesteps %d maxatom %d maxseries %d' % (self.correlArrayLength, 1000, len(group)))
for j, contact in enumerate(group):
String.append('enter n%03d bond bynum %d bynum %d geometry' % (j, contact.i.atomNum, contact.j.atomNum))
String.append('traj firstu 10 nunit 1 begin %d stop %d skip %d select all end' % (self.correlStart, self.correlStop, self.correlSkip))
String.append('')
for j, contact in enumerate(group):
String.append('write n%03d card unit %03d' % (j, j+100))
String.append('* Contact %02d%02d: between cgAtoms %s and %s' % (i, j, contact.i.addr, contact.j.addr))
String.append('* Native Contact - Interaction between %s and %s' % (contact.i.prmString, contact.j.prmString))
String.append('*')
String.append('')
String.append('end')
String.append('')
String.append('rewind unit 10')
String.append('')
String.append('stop')
String = '\n'.join(String)
# Write file
self.inpFilename = '%s/natq%02d.inp' % (self.inpPathname, i)
print '%s' % self.inpFilename
write_to = open(self.inpFilename, 'w')
write_to.write(String)
write_to.close()
def run_correlInput(self):
#
for i, group in enumerate(grouper(self.nativeContacts, 100)):
self.inpFilename = '%s/natq%02d.inp' % (self.inpPathname, i)
self.outFilename = '%s/natq%02d.out' % (self.outPathname, i)
#
try:
os.remove(self.outFilename)
except OSError:
pass
if not os.path.exists(self.inpFilename):
raise IOError("No such file or directory: '%s'" % self.inpFilename)
os.system('%s < %s > %s' % (self.charmmBin, self.inpFilename, self.outFilename))
class Contacts(BaseAnalysis):
"""
DOCME
"""
def __init__(self, pdbFilename=None, **kwargs):
super(Contacts, self).__init__(pdbFilename, **kwargs)
# kwargs
self.nativeRad = kwargs.get('nativerad', 8.)
# Main
if pdbFilename is not None:
self.aa = PDBFile(self.pdbFilename)[0]
self.aa.parse()
self.cg = KTGo(self.aa)
self.contacts = self.get_contacts()
@Property
def data():
doc =\
"""
"""
def fget(self):
filename = '%s/contacts.pickle' % self.anlPathname
try:
self._data = pickle.load(open(filename))
print 'found pickled data in: %s' % filename
return self._data
except IOError:
print 'processing data in: %s' % self.anlPathname
self._data = self.build_contactMatrix()
pickle.dump(self._data, open(filename, 'w'))
return self._data
def fset(self, value):
assert isinstance(value, np.ndarray)
self._data = value
def fdel(self):
filename = '%s/contacts.pickle' % self.anlPathname
del self._data
try:
os.remove(filename)
except IOError:
pass
return locals()
@Property
def inpPathname():
doc =\
"""
"""
def fget(self):
return self._inpPathname
def fset(self, value):
self._inpPathname = self.expandPath(value)
return locals()
@Property
def nativeRad():
doc =\
"""
The cutoff distance between two CG sidechain atom objects which is used
to determine if their parent residues are in 'contact'.
"""
def fget(self):
return self._nativeRad
def fset(self, value):
if value <= 1:
raise ValueError("The native contact radius must be at least 1.")
self._nativeRad = float(value)
return locals()
def get_contacts(self):
"""
Returns a complete :class:`list` of :class:`Bond` objects, each bond
object represents a unique sidechain-sidechain pair. Additionally,
and most importantly, each :class:`Bond` object has been assigned the
additional ``native`` attribute, to ``True`` or ``False``. This value
is true, if the corresponding all-atom representation of the residue
sidechains have any heavy atoms within 4.5 Angstrom, as determined by
:meth:`KTGo.get_nativeSCSC`.
"""
iterator = [ atom for atom in self.cg if atom.atomType == 's' ]
tmp = [ Bond(atom_i, atom_j) for i, atom_i in enumerate(iterator) for
j, atom_j in enumerate(iterator) if i < j ]
nativeContacts = set(self.cg.get_nativeSCSC())
for contact in tmp:
if contact in nativeContacts:
contact.native = True
else:
contact.native = False
return tmp
def build_contactMatrix(self):
"""
Builds a 2D numpy array from available correl output files. The first
index references the contact number, the second index references the
correl time array value (which in turn corresponds to nstep of the
original dynamics in quanta of ``correlSkip``. This method can take
awhile because of disk IO.
"""
print "Building full contact matrix."
rowLength = (self.correlStop - self.correlStart) / self.correlSkip + 1
tmp = []
for i in xrange(len(self.contacts)):
try:
tmp.append(load_correlOutput('%s/contact%04d.anl' % (self.anlPathname, i)))
except IOError:
tmp.append(np.zeros(rowLength))
print "Can't find correl output for contact number: %04d." % i
return np.array(tmp, dtype=np.float64)
def get_nativeContactMatrix(self):
"""
Projects out elements of the contactMatrix which are not derived from
'native' contacts.
"""
nativeIndex = [ i for i, taco in enumerate(self.contacts) if taco.native ]
return self.data[nativeIndex]
def get_natQofT(self):
"""
"""
nativeContacts = self.get_nativeContactMatrix()
tmp = np.zeros(nativeContacts.shape)
tmp[nativeContacts <= self.nativeRad] = 1
tmp[nativeContacts > self.nativeRad] = 0
return tmp.mean(axis=0)
# Charmm input creation
def do_correl(self, **kwargs):
self.write_correlInput(**kwargs)
self.run_correlInput()
def write_correlInput(self, **kwargs):
kwargs = lowerKeys(kwargs)
header = kwargs.get('header', [])
#
if self.anlPathname is None:
raise IOError("No directory specified.")
mkdir(self.anlPathname)
#
for i, group in enumerate(grouper(self.contacts, 100)):
String = self.get_correlInputHeader(header)
String.append('! anl :: write')
for j, contact in enumerate(group):
String.append('open unit %03d write card name %s/contact%02d%02d.anl' % (j+100, self.anlPathname, i, j))
String.append('')
String.append('correl maxtimesteps %d maxatom %d maxseries %d' % (self.correlArrayLength, 1000, len(group)))
for j, contact in enumerate(group):
String.append('enter n%03d bond bynum %d bynum %d geometry' % (j, contact.i.atomNum, contact.j.atomNum))
String.append('traj firstu 10 nunit 1 begin %d stop %d skip %d select all end' % (self.correlStart, self.correlStop, self.correlSkip))
String.append('')
for j, contact in enumerate(group):
String.append('write n%03d card unit %03d' % (j, j+100))
String.append('* Contact %02d%02d: between cgAtoms %s and %s' % (i, j, contact.i.addr, contact.j.addr))
if contact.native:
String.append('* Native Contact - Interaction between %s and %s' % (contact.i.prmString, contact.j.prmString))
String.append('*')
String.append('')
String.append('end')
String.append('')
String.append('rewind unit 10')
String.append('')
String.append('stop')
String = '\n'.join(String)
# Write file
self.inpFilename = '%s/contact%02d.inp' % (self.inpPathname, i)
write_to = open(self.inpFilename, 'w')
write_to.write(String)
write_to.close()
def run_correlInput(self):
#
for i, group in enumerate(grouper(self.contacts, 100)):
self.inpFilename = '%s/contact%02d.inp' % (self.inpPathname, i)
self.outFilename = '%s/contact%02d.out' % (self.outPathname, i)
#
try:
os.remove(self.outFilename)
except OSError:
pass
if not os.path.exists(self.inpFilename):
raise IOError("No such file or directory: '%s'" % self.inpFilename)
os.system('%s < %s > %s' % (self.charmmBin, self.inpFilename, self.outFilename))