def data3DList(c, key='fnac_10', inverse=0, rm=range(1, 12), soln=512):
"""
Create an matrix: len(rec_model) * len((lig_model) * solutions
containing the values of the info dic with given key.
c - ComplexList
"""
rm = range(1, max(c.valuesOf('model1')) + 1)
lm = range(1, max(c.valuesOf('model2')) + 1)
matrix = zeros((len(rm), len(lm), soln), 'f')
try:
for r in rm:
rl = c.filter('model1', r)
for l in lm:
cl = rl.filter('model2', l)
if inverse:
matrix[r - 1][l - 1] = (
1. / array(cl.valuesOf(key, default=0))).tolist()
else:
matrix[r - 1][l - 1] = cl.valuesOf(key, default=0)
except ValueError, why:
try:
lenM = len(matrix[r - 1][l - 1])
lenV = len(cl.valuesOf(key, default=0))
except:
lenM = lenV = 0
s = '%i : %i len(matrix)=%i <> len(values)=%i' % (r, l, lenM, lenV)
EHandler.error('Cannot extract fnac data. ' + s)
def concat( self, *traj ):
"""
Concatenate this with other trajectories. The ref model of the
new Trajectory is a 'semi-deep' copy of this trajectorie's model.
(see L{PDBModel.take()} )::
concat( traj [, traj2, traj3, ..] ) -> Trajectory
@param traj: one or more Trajectory with identical atoms as this one
@type traj: Trajectories
@return: concatenated trajecties
@rtype: Trajectory
"""
if len( traj ) == 0:
return self
r = self.__class__()
r.frames = N.concatenate( (self.frames, traj[0].frames), 0 )
r.setRef( self.ref.clone())
if self.frameNames and traj[0].frameNames:
r.frameNames = self.frameNames + traj[0].frameNames
try:
if self.pc is not None and traj[0].pc is not None:
r.pc['p'] = N.concatenate( (self.pc['p'], traj[0].pc['p']),0)
r.pc['u'] = N.concatenate( (self.pc['u'], traj[0].pc['u']),0)
except TypeError, why:
EHandler.error('cannot concat PC '+str(why) )
def data3DList( c, key='fnac_10', inverse=0,
rm=range(1,12), soln=512 ):
"""
Create an matrix: len(rec_model) * len((lig_model) * solutions
containing the values of the info dic with given key.
c - ComplexList
"""
rm = range( 1, max(c.valuesOf( 'model1'))+1 )
lm = range( 1, max(c.valuesOf( 'model2'))+1 )
matrix = zeros( ( len( rm ), len( lm ), soln ), 'f' )
try:
for r in rm:
rl = c.filter( 'model1', r )
for l in lm:
cl = rl.filter( 'model2', l )
if inverse:
matrix[r-1][l-1] = (1./array(
cl.valuesOf( key, default=0 ))).tolist()
else:
matrix[r-1][l-1] = cl.valuesOf( key, default=0 )
except ValueError, why:
try:
lenM = len( matrix[r-1][l-1] )
lenV = len( cl.valuesOf( key, default=0 ) )
except:
lenM = lenV = 0
s = '%i : %i len(matrix)=%i <> len(values)=%i' % (r, l, lenM, lenV)
EHandler.error('Cannot extract fnac data. '+ s )
def concat(self, *traj):
"""
Concatenate this with other trajectories. The ref model of the
new Trajectory is a 'semi-deep' copy of this trajectorie's model.
(see L{PDBModel.take()} )::
concat( traj [, traj2, traj3, ..] ) -> Trajectory
@param traj: one or more Trajectory with identical atoms as this one
@type traj: Trajectories
@return: concatenated trajecties
@rtype: Trajectory
"""
if len(traj) == 0:
return self
r = self.__class__()
r.frames = N.concatenate((self.frames, traj[0].frames), 0)
r.setRef(self.ref.clone())
if self.frameNames and traj[0].frameNames:
r.frameNames = self.frameNames + traj[0].frameNames
try:
if self.pc is not None and traj[0].pc is not None:
r.pc['p'] = N.concatenate((self.pc['p'], traj[0].pc['p']), 0)
r.pc['u'] = N.concatenate((self.pc['u'], traj[0].pc['u']), 0)
except TypeError, why:
EHandler.error('cannot concat PC ' + str(why))
def add(self, str):
"""
Add String str and line break to xplor input file.
@param str: string to add to file
@type str: str
"""
try:
self.fgenerate.write(str + '\n')
except (IOError):
EHandler.error(
"XPlorInput.append(): Error adding str to xplor input file.")
def add(self, str):
"""
Add String str and line break to xplor input file.
@param str: string to add to file
@type str: str
"""
try:
self.fgenerate.write(str + '\n')
except (IOError):
EHandler.error(
"XPlorInput.append(): Error adding str to xplor input file.")
def writePdb( self, index, fname):
"""
Write (possibly transformed) coordinates back to pdb.
@param index: frame index in trajectory
@type index: int
@param fname: name of new file
@type fname: str
"""
try:
self.getPDBModel( index ).writePdb( fname )
except:
EHandler.error('Error writing %s.' % fname)
def writePdb(self, index, fname):
"""
Write (possibly transformed) coordinates back to pdb.
@param index: frame index in trajectory
@type index: int
@param fname: name of new file
@type fname: str
"""
try:
self.getPDBModel(index).writePdb(fname)
except:
EHandler.error('Error writing %s.' % fname)
## note 1: If there are more than approximately 50 sequences overall
## t_coffe will eat all the memory and the job will not finish
## This should be fixed in more recent versions of T-Coffee
## (v > 3.2) where T-Coffee, according to the manual "switches
## to a heuristic mode, named DPA, where DPA stands for Double
## Progressive Alignment."
## note 2: If there is only one template structure step 2 of T-coffee
## will not work. Solution, skip the structural alignment if
## only one template structure is provided.
## note 3: In quite som cases the sequence retrieved from the nrpdb
## sequence database is different from the sequence extracted
## from the coordinates in the pdb-file. This will sometimes
## cause t-coffee to terminate with an error (two sequences
## with the same name but with different sequences). Temporary
## solution: Choose another structure from the same cluster
## as the troublemaker.
try:
a = Aligner(outFolder, log, verbose=1, sap=sap)
a.align_for_modeller_inp()
a.go(host)
except:
EHandler.error('Error while building alingnments.')
print "\nalign.py -? or align.py -help for help screen"
# searcher.localPSIBlast( target, seq_db, e=0.1, alignments=1000)
## local Blast
searcher.localBlast( f_target, seq_db, 'blastp', alignments=500, e=0.0001 )
## cluster blast results. Defaults: simCut=1.75, lenCut=0.9, ncpu=1
## expects all.fasta
# searcher.clusterFastaIterative( )
searcher.clusterFasta()
searcher.writeFastaClustered()
tools.flushPrint('Done.\n')
except:
EHandler.error( 'Error while searching for homologues.')
###############
## TemplateSearcher
##
## Find modelling templates, blasting the target sequence against "tmp_db"
## Cluster the sequences and download the pdbs to templates/all
## input: target.fasta
##
## output: templates/blast.out
## templates/all.fasta
## templates/cluster_result.out
## templates/nr.fasta (input for Aligner)
## templates/all/*.pdb
if '?' in options or 'help' in options:
_use(defaultOptions())
log = None
if options['log']:
log = LogFile(outFolder + '/' + options['log'], 'a')
###################
## TemplateCleaner
##
## Prepare pdb files in templates/nr for T-coffee and modeller
## (replace nonstandard residues, remove hydrogens,
# remove atoms with nultiple configurations, etc.)
## input: templates/nr/*.pdb
## templates/nr/chain_index.txt
##
## output: templates/t_coffee/*.alpha (input for Alignar)
## templates/modeller/*.pdb (input for Modeller)
try:
cleaner = TemplateCleaner(outFolder, log)
inp_dic = modUtils.parse_tabbed_file(chIndex)
cleaner.process_all(inp_dic)
except:
EHandler.error('Error while cleaning templates')
## t_coffe will eat all the memory and the job will not finish
## This should be fixed in more recent versions of T-Coffee
## (v > 3.2) where T-Coffee, according to the manual "switches
## to a heuristic mode, named DPA, where DPA stands for Double
## Progressive Alignment."
## note 2: If there is only one template structure step 2 of T-coffee
## will not work. Solution, skip the structural alignment if
## only one template structure is provided.
## note 3: In quite som cases the sequence retrieved from the nrpdb
## sequence database is different from the sequence extracted
## from the coordinates in the pdb-file. This will sometimes
## cause t-coffee to terminate with an error (two sequences
## with the same name but with different sequences). Temporary
## solution: Choose another structure from the same cluster
## as the troublemaker.
try:
a = Aligner( outFolder, log, verbose=1, sap=sap )
a.align_for_modeller_inp()
a.go(host)
except:
EHandler.error( 'Error while building alingnments.')
print "\nalign.py -? or align.py -help for help screen"
##
## output: modeller/modeller.log
## /*.B9999000?? <- models
try:
if options['verbose'] > 0:
print "\n"+\
"Type model.py -? or model.py -help for a full list of options!"
m8 = M(**options)
if not 'dry' in options:
r = m8.run() ## comment out for testing
except:
EHandler.error('Error while modelling.')
#####################
## Show output
## show result in PyMol
if options.has_key('s'):
names = []
## fit backbone of all models to average until convergence
models = glob.glob('%s/target.B*' % (m8.outFolder + m8.F_RESULT_FOLDER))
traj = Trajectory(models)
traj.blockFit2ref(mask=traj[0].maskBB())
## calculate and print rmsd matrix
rmsHeavy = traj.pairwiseRmsd()
if '?' in options or 'help' in options:
_use( defaultOptions() )
log = None
if options['log']:
log = LogFile( outFolder + '/' + options['log'], 'a' )
###################
## TemplateCleaner
##
## Prepare pdb files in templates/nr for T-coffee and modeller
## (replace nonstandard residues, remove hydrogens,
# remove atoms with nultiple configurations, etc.)
## input: templates/nr/*.pdb
## templates/nr/chain_index.txt
##
## output: templates/t_coffee/*.alpha (input for Alignar)
## templates/modeller/*.pdb (input for Modeller)
try:
cleaner = TemplateCleaner( outFolder, log )
inp_dic = modUtils.parse_tabbed_file( chIndex )
cleaner.process_all( inp_dic )
except:
EHandler.error( 'Error while cleaning templates')
# searcher.localPSIBlast( target, seq_db, e=0.1, alignments=1000)
## local Blast
searcher.localBlast(f_target, seq_db, 'blastp', alignments=500, e=0.0001)
## cluster blast results. Defaults: simCut=1.75, lenCut=0.9, ncpu=1
## expects all.fasta
# searcher.clusterFastaIterative( )
searcher.clusterFasta()
searcher.writeFastaClustered()
tools.flushPrint('Done.\n')
except:
EHandler.error('Error while searching for homologues.')
###############
## TemplateSearcher
##
## Find modelling templates, blasting the target sequence against "tmp_db"
## Cluster the sequences and download the pdbs to templates/all
## input: target.fasta
##
## output: templates/blast.out
## templates/all.fasta
## templates/cluster_result.out
## templates/nr.fasta (input for Aligner)
## templates/all/*.pdb
## templates/nr/chain_index.txt (input for TemplateCleaner)
def concatEnsembles(self, *traj):
"""
Concatenate this with other trajectories in a zig zac manner,
resulting in an ensembleTraj with additional members.
The ref model of the new Trajectory is a 'semi-deep' copy of this
trajectorie's model.(see L{PDBModel.take()} )::
concat( traj [, traj2, traj3, ..] ) -> Trajectory
@param traj: with identical atoms as this one
@type traj: one or more EnsembleTrajectory
@todo: fix so that pc, and profiles are not lost
"""
if len(traj) == 0:
return self
r = self.__class__(n_members=self.n_members + traj[0].n_members)
min_members = min(self.n_members, traj[0].n_members)
min_frames = min(self.lenFrames(), traj[0].lenFrames())
steps = self.lenFrames()/self.n_members + \
traj[0].lenFrames()/traj[0].n_members
def __everyOther(traj_0, traj_1, list_0, list_1, minMembers, minFrames,
loops):
result = []
for j in range(0, minMembers / 2):
for i in range(j * loops,
j * loops + minFrames * 2 / minMembers):
result += [list_0[i]]
result += [list_1[i]]
while i < j * traj_0.n_members:
result += [list_0[i]]
while i < j * traj_1.n_members:
result += [list_1[i]]
return result
frames = __everyOther(self, traj[0], self.frames, traj[0].frames,
min_members, min_frames, steps)
r.frames = N0.array(frames)
r.setRef(self.ref.clone())
if self.frameNames and traj[0].frameNames:
r.frameNames = __everyOther(self, traj[0], self.frameNames,
traj[0].frameNames, min_members,
min_frames, steps)
try:
# NOT TESTED!!
if self.pc and traj[0].pc:
r.pc['p'] = __everyOther(self, traj[0], self.pc['p'],
traj[0].pc['p'], min_members, steps)
r.pc['u'] = __everyOther(self, traj[0], self.pc['u'],
traj[0].pc['u'], min_members, steps)
# r.pc['p'] = N0.concatenate( (self.pc['p'], traj[0].pc['p']),0)
# r.pc['u'] = N0.concatenate( (self.pc['u'], traj[0].pc['u']),0)
except TypeError, why:
EHandler.error('cannot concat PC ' + str(why))
def concatEnsembles( self, *traj ):
"""
Concatenate this with other trajectories in a zig zac manner,
resulting in an ensembleTraj with additional members.
The ref model of the new Trajectory is a 'semi-deep' copy of this
trajectorie's model.(see L{PDBModel.take()} )::
concat( traj [, traj2, traj3, ..] ) -> Trajectory
@param traj: with identical atoms as this one
@type traj: one or more EnsembleTrajectory
@todo: fix so that pc, and profiles are not lost
"""
if len( traj ) == 0:
return self
r = self.__class__( n_members = self.n_members + traj[0].n_members )
min_members = min( self.n_members, traj[0].n_members )
min_frames = min( self.lenFrames(), traj[0].lenFrames() )
steps = self.lenFrames()/self.n_members + \
traj[0].lenFrames()/traj[0].n_members
def __everyOther( traj_0, traj_1, list_0, list_1, minMembers,
minFrames, loops ):
result = []
for j in range( 0, minMembers/2 ):
for i in range( j*loops , j*loops + minFrames*2/minMembers ):
result += [ list_0[i] ]
result += [ list_1[i] ]
while i < j*traj_0.n_members:
result += [ list_0[i] ]
while i < j*traj_1.n_members:
result += [ list_1[i] ]
return result
frames = __everyOther( self, traj[0], self.frames,
traj[0].frames, min_members,
min_frames, steps )
r.frames = N.array(frames)
r.setRef( self.ref.clone())
if self.frameNames and traj[0].frameNames:
r.frameNames = __everyOther( self, traj[0], self.frameNames,
traj[0].frameNames, min_members,
min_frames, steps )
try:
# NOT TESTED!!
if self.pc and traj[0].pc:
r.pc['p'] = __everyOther( self, traj[0], self.pc['p'],
traj[0].pc['p'], min_members, steps )
r.pc['u'] = __everyOther( self, traj[0], self.pc['u'],
traj[0].pc['u'], min_members, steps )
# r.pc['p'] = N.concatenate( (self.pc['p'], traj[0].pc['p']),0)
# r.pc['u'] = N.concatenate( (self.pc['u'], traj[0].pc['u']),0)
except TypeError, why:
EHandler.error('cannot concat PC '+str(why) )