def calcGNMfeatures(self, chain='all', env='chain', GNM_PRS=True):
"""Computes GNM-based features.
:arg chain: chain identifier
:type chain: str
:arg env: environment model, i.e. ``'chain'``, ``'reduced'`` or
``'sliced'``
:type env: str
:arg GNM_PRS: whether or not to compute features based on Perturbation
Response Scanning analysis
:type GNM_PRS: bool
"""
assert env in ['chain', 'reduced', 'sliced']
assert type(GNM_PRS) is bool
# list of features to be computed
features = ['GNM_MSF-'+env]
if GNM_PRS:
features += ['GNM_effectiveness-'+env, 'GNM_sensitivity-'+env]
# compute features (if not precomputed)
if chain == 'all':
chain_list = self.chids
else:
chain_list = [chain, ]
for chID in chain_list:
d = self.feats[chID]
if all([f in d for f in features]):
continue
try:
gnm = self.calcGNM(chID, env=env)
except Exception as e:
if (isinstance(e, MemoryError)):
msg = 'MemoryError'
else:
msg = str(e)
for f in features:
d[f] = msg
LOGGER.warn(msg)
continue
key_msf = 'GNM_MSF-' + env
if key_msf not in d:
try:
d[key_msf] = calcSqFlucts(gnm)
except Exception as e:
msg = str(e)
d[key_msf] = msg
LOGGER.warn(msg)
key_eff = 'GNM_effectiveness-' + env
if key_eff in features and key_eff not in d:
key_sns = 'GNM_sensitivity-' + env
try:
prs_mtrx, eff, sns = calcPerturbResponse(gnm)
d[key_eff] = eff
d[key_sns] = sns
except Exception as e:
msg = str(e)
d[key_eff] = msg
d[key_sns] = msg
LOGGER.warn(msg)
return
def _do_translation(self):
weights = 1.0 / prody.calcSqFlucts(self._anm)
#x_mean = numpy.mean(self._x, axis=0)
#y_mean = numpy.mean(self._y, axis=0)
x_mean = calc_average_coords(self._x, weights)
y_mean = calc_average_coords(self._y, weights)
self._x = self._x - x_mean
self._y = self._y - y_mean
self._x_mean = x_mean
self._y_mean = y_mean
def _do_translation(self):
weights = 1.0 / prody.calcSqFlucts(self._anm)
#x_mean = numpy.mean(self._x, axis=0)
#y_mean = numpy.mean(self._y, axis=0)
x_mean = calc_average_coords(self._x, weights)
y_mean = calc_average_coords(self._y, weights)
self._x = self._x - x_mean
self._y = self._y - y_mean
self._x_mean = x_mean
self._y_mean = y_mean
def calcGNMfeatures(self, chain='all', env='chain', GNM_PRS=True):
assert env in ['chain', 'reduced', 'sliced']
assert type(GNM_PRS) is bool
# list of features to be computed
features = ['GNM_MSF-' + env]
if GNM_PRS:
features += ['GNM_effectiveness-' + env, 'GNM_sensitivity-' + env]
# compute features (if not precomputed)
if chain == 'all':
chain_list = self.chids
else:
chain_list = [
chain,
]
for chID in chain_list:
d = self.feats[chID]
if all([f in d for f in features]):
continue
try:
gnm = self.calcGNM(chID, env=env)
except Exception as e:
if (isinstance(e, MemoryError)):
msg = 'MemoryError'
else:
msg = str(e)
for f in features:
d[f] = msg
LOGGER.warn(msg)
continue
key_msf = 'GNM_MSF-' + env
if key_msf not in d:
try:
d[key_msf] = calcSqFlucts(gnm)
except Exception as e:
msg = str(e)
d[key_msf] = msg
LOGGER.warn(msg)
key_eff = 'GNM_effectiveness-' + env
if key_eff in features and key_eff not in d:
key_sns = 'GNM_sensitivity-' + env
try:
prs_mtrx, eff, sns = calcPerturbResponse(gnm)
d[key_eff] = eff
d[key_sns] = sns
except Exception as e:
msg = str(e)
d[key_eff] = msg
d[key_sns] = msg
LOGGER.warn(msg)
return
def get_enm_fluctuations(enm, n_modes=6):
"""
Get squared fluctuations of each residue according to an elastic network model
Parameters
----------
enm
pd.dynamics.anm.ANM or pd.dynamics.gnm.GNM object
n_modes
number of ENM modes to consider
Returns
-------
array of squared fluctuations per residue
"""
return pd.calcSqFlucts(enm[:n_modes])
def get_pca_fluctuations(ensemble, limit=3):
"""
Get squared fluctuations of each residue according to a PCA on the ensemble
Parameters
----------
ensemble
pd.PDBEnsemble object
limit
number of PCA modes to consider
Returns
-------
array of squared fluctuations per aligned residue
"""
pca = pd.PCA()
pca.buildCovariance(ensemble)
pca.calcModes()
return pd.calcSqFlucts(pca[:limit])
def prody_anm(pdb, **kwargs):
"""Perform ANM calculations for *pdb*.
"""
for key in DEFAULTS:
if not key in kwargs:
kwargs[key] = DEFAULTS[key]
from os.path import isdir, join
outdir = kwargs.get('outdir')
if not isdir(outdir):
raise IOError('{0} is not a valid path'.format(repr(outdir)))
import numpy as np
import prody
LOGGER = prody.LOGGER
selstr = kwargs.get('select')
prefix = kwargs.get('prefix')
cutoff = kwargs.get('cutoff')
gamma = kwargs.get('gamma')
nmodes = kwargs.get('nmodes')
selstr = kwargs.get('select')
model = kwargs.get('model')
pdb = prody.parsePDB(pdb, model=model)
if prefix == '_anm':
prefix = pdb.getTitle() + '_anm'
select = pdb.select(selstr)
if select is None:
LOGGER.warn('Selection {0} did not match any atoms.'.format(
repr(selstr)))
return
LOGGER.info('{0} atoms will be used for ANM calculations.'.format(
len(select)))
anm = prody.ANM(pdb.getTitle())
anm.buildHessian(select, cutoff, gamma)
anm.calcModes(nmodes)
LOGGER.info('Writing numerical output.')
if kwargs.get('outnpz'):
prody.saveModel(anm, join(outdir, prefix))
prody.writeNMD(join(outdir, prefix + '.nmd'), anm, select)
extend = kwargs.get('extend')
if extend:
if extend == 'all':
extended = prody.extendModel(anm, select, pdb)
else:
extended = prody.extendModel(anm, select, select | pdb.bb)
prody.writeNMD(join(outdir, prefix + '_extended_' + extend + '.nmd'),
*extended)
outall = kwargs.get('outall')
delim = kwargs.get('numdelim')
ext = kwargs.get('numext')
format = kwargs.get('numformat')
if outall or kwargs.get('outeig'):
prody.writeArray(join(outdir, prefix + '_evectors' + ext),
anm.getArray(),
delimiter=delim,
format=format)
prody.writeArray(join(outdir, prefix + '_evalues' + ext),
anm.getEigvals(),
delimiter=delim,
format=format)
if outall or kwargs.get('outbeta'):
from prody.utilities import openFile
fout = openFile(prefix + '_beta.txt', 'w', folder=outdir)
fout.write(
'{0[0]:1s} {0[1]:4s} {0[2]:4s} {0[3]:5s} {0[4]:5s}\n'.format(
['C', 'RES', '####', 'Exp.', 'The.']))
for data in zip(select.getChids(), select.getResnames(),
select.getResnums(), select.getBetas(),
prody.calcTempFactors(anm, select)):
fout.write(
'{0[0]:1s} {0[1]:4s} {0[2]:4d} {0[3]:5.2f} {0[4]:5.2f}\n'.
format(data))
fout.close()
if outall or kwargs.get('outcov'):
prody.writeArray(join(outdir, prefix + '_covariance' + ext),
anm.getCovariance(),
delimiter=delim,
format=format)
if outall or kwargs.get('outcc') or kwargs.get('outhm'):
cc = prody.calcCrossCorr(anm)
if outall or kwargs.get('outcc'):
prody.writeArray(join(outdir,
prefix + '_cross-correlations' + ext),
cc,
delimiter=delim,
format=format)
if outall or kwargs.get('outhm'):
prody.writeHeatmap(join(outdir, prefix + '_cross-correlations.hm'),
cc,
resnum=select.getResnums(),
xlabel='Residue',
ylabel='Residue',
title=anm.getTitle() + ' cross-correlations')
if outall or kwargs.get('hessian'):
prody.writeArray(join(outdir, prefix + '_hessian' + ext),
anm.getHessian(),
delimiter=delim,
format=format)
if outall or kwargs.get('kirchhoff'):
prody.writeArray(join(outdir, prefix + '_kirchhoff' + ext),
anm.getKirchhoff(),
delimiter=delim,
format=format)
if outall or kwargs.get('outsf'):
prody.writeArray(join(outdir, prefix + '_sqflucts' + ext),
prody.calcSqFlucts(anm),
delimiter=delim,
format=format)
figall = kwargs.get('figall')
cc = kwargs.get('figcc')
sf = kwargs.get('figsf')
bf = kwargs.get('figbeta')
cm = kwargs.get('figcmap')
if figall or cc or sf or bf or cm:
try:
import matplotlib.pyplot as plt
except ImportError:
LOGGER.warning('Matplotlib could not be imported. '
'Figures are not saved.')
else:
prody.SETTINGS['auto_show'] = False
LOGGER.info('Saving graphical output.')
format = kwargs.get('figformat')
width = kwargs.get('figwidth')
height = kwargs.get('figheight')
dpi = kwargs.get('figdpi')
format = format.lower()
if figall or cc:
plt.figure(figsize=(width, height))
prody.showCrossCorr(anm)
plt.savefig(join(outdir, prefix + '_cc.' + format),
dpi=dpi,
format=format)
plt.close('all')
if figall or cm:
plt.figure(figsize=(width, height))
prody.showContactMap(anm)
plt.savefig(join(outdir, prefix + '_cm.' + format),
dpi=dpi,
format=format)
plt.close('all')
if figall or sf:
plt.figure(figsize=(width, height))
prody.showSqFlucts(anm)
plt.savefig(join(outdir, prefix + '_sf.' + format),
dpi=dpi,
format=format)
plt.close('all')
if figall or bf:
plt.figure(figsize=(width, height))
bexp = select.getBetas()
bcal = prody.calcTempFactors(anm, select)
plt.plot(bexp, label='Experimental')
plt.plot(bcal,
label=('Theoretical (R={0:.2f})'.format(
np.corrcoef(bcal, bexp)[0, 1])))
plt.legend(prop={'size': 10})
plt.xlabel('Node index')
plt.ylabel('Experimental B-factors')
plt.title(pdb.getTitle() + ' B-factors')
plt.savefig(join(outdir, prefix + '_bf.' + format),
dpi=dpi,
format=format)
plt.close('all')
def prody_anm(opt):
"""Perform ANM calculations based on command line arguments."""
outdir = opt.outdir
if not os.path.isdir(outdir):
opt.subparser.error('{0:s} is not a valid path'.format(outdir))
import numpy as np
import prody
LOGGER = prody.LOGGER
pdb = opt.pdb
prefix = opt.prefix
cutoff, gamma = opt.cutoff, opt.gamma,
nmodes, selstr, model = opt.nmodes, opt.select, opt.model
pdb = prody.parsePDB(pdb, model=model)
if prefix == '_anm':
prefix = pdb.getTitle() + '_anm'
select = pdb.select(selstr)
if select is None:
opt.subparser('Selection "{0:s}" do not match any atoms.'
.format(selstr))
LOGGER.info('{0:d} atoms will be used for ANM calculations.'
.format(len(select)))
anm = prody.ANM(pdb.getTitle())
anm.buildHessian(select, cutoff, gamma)
anm.calcModes(nmodes)
LOGGER.info('Writing numerical output.')
if opt.npz:
prody.saveModel(anm)
prody.writeNMD(os.path.join(outdir, prefix + '.nmd'), anm, select)
outall = opt.all
delim, ext, format = opt.delim, opt.ext, opt.numformat
if outall or opt.eigen:
prody.writeArray(os.path.join(outdir, prefix + '_evectors'+ext),
anm.getArray(), delimiter=delim, format=format)
prody.writeArray(os.path.join(outdir, prefix + '_evalues'+ext),
anm.getEigenvalues(), delimiter=delim, format=format)
if outall or opt.beta:
fout = prody.openFile(prefix + '_beta.txt', 'w', folder=outdir)
fout.write('{0[0]:1s} {0[1]:4s} {0[2]:4s} {0[3]:5s} {0[4]:5s}\n'
.format(['C', 'RES', '####', 'Exp.', 'The.']))
for data in zip(select.getChids(), select.getResnames(),
select.getResnums(), select.getBetas(),
prody.calcTempFactors(anm, select)):
fout.write('{0[0]:1s} {0[1]:4s} {0[2]:4d} {0[3]:5.2f} {0[4]:5.2f}\n'
.format(data))
fout.close()
if outall or opt.covar:
prody.writeArray(os.path.join(outdir, prefix + '_covariance'+ext),
anm.getCovariance(), delimiter=delim, format=format)
if outall or opt.ccorr:
prody.writeArray(os.path.join(outdir, prefix + '_cross-correlations'
+ ext),
prody.calcCrossCorr(anm), delimiter=delim,
format=format)
if outall or opt.hessian:
prody.writeArray(os.path.join(outdir, prefix + '_hessian'+ext),
anm.getHessian(), delimiter=delim, format=format)
if outall or opt.kirchhoff:
prody.writeArray(os.path.join(outdir, prefix + '_kirchhoff'+ext),
anm.getKirchhoff(), delimiter=delim, format=format)
if outall or opt.sqflucts:
prody.writeArray(os.path.join(outdir, prefix + '_sqflucts'+ext),
prody.calcSqFlucts(anm), delimiter=delim,
format=format)
figall, cc, sf, bf, cm = opt.figures, opt.cc, opt.sf, opt.bf, opt.cm
if figall or cc or sf or bf or cm:
try:
import matplotlib.pyplot as plt
except ImportError:
LOGGER.warning('Matplotlib could not be imported. '
'Figures are not saved.')
else:
LOGGER.info('Saving graphical output.')
format, width, height, dpi = \
opt.figformat, opt.width, opt.height, opt.dpi
format = format.lower()
if figall or cc:
plt.figure(figsize=(width, height))
prody.showCrossCorr(anm)
plt.savefig(os.path.join(outdir, prefix + '_cc.'+format),
dpi=dpi, format=format)
plt.close('all')
if figall or cm:
plt.figure(figsize=(width, height))
prody.showContactMap(anm)
plt.savefig(os.path.join(outdir, prefix + '_cm.'+format),
dpi=dpi, format=format)
plt.close('all')
if figall or sf:
plt.figure(figsize=(width, height))
prody.showSqFlucts(anm)
plt.savefig(os.path.join(outdir, prefix + '_sf.'+format),
dpi=dpi, format=format)
plt.close('all')
if figall or bf:
plt.figure(figsize=(width, height))
bexp = select.getBetas()
bcal = prody.calcTempFactors(anm, select)
plt.plot(bexp, label='Experimental')
plt.plot(bcal, label=('Theoretical (R={0:.2f})'
.format(np.corrcoef(bcal, bexp)[0,1])))
plt.legend(prop={'size': 10})
plt.xlabel('Node index')
plt.ylabel('Experimental B-factors')
plt.title(pdb.getTitle() + ' B-factors')
plt.savefig(os.path.join(outdir, prefix + '_bf.'+format),
dpi=dpi, format=format)
plt.close('all')
def corepagecalculation(pdbfilename, selatom, noma1, nummodes, gamcut, cut1, gam2, cut2, showresults, smodes, snmd, smodel, scollec, massnomass, sample1, modeens, confens, rmsdens, traverse1, modetra, steptra, rmsdtra, modelnumber, caanm, cagnm, nohanm, nohgnm, allanm, allgnm, bbanm, bbgnm, scanm, scgnm, nmdfolder, modesfolder, collectivityfolder, modelnewname, nmdnewname, modesnewname, modesendname, collectivitynewname, collectivityendname, samplenewname, traversenewname, crosscorr=0, corrfolder='', corrname='', corrend='', compmode01='7', compmode02='15', sqflucts=0, sqfluctsfolder='', sqfluctsname='', sqfluctsend='', separatevar1='0', temfac=0, temfacfolder='', temfacname='', temfacend='', fracovar=0, fraconame='', fracoend='', ovlap=0, ovlapfold='', ovlapname='', ovlapend='', ovlaptab=0, ovlaptabname='', ovlaptabend='', comppdbfilename=''):
# modelnumber
import prody
import time
import os
import Tkinter
root=Tkinter.Tk()
root.title('Info')
onlypage=Tkinter.Frame(root)
onlypage.pack(side='top')
Tkinter.Label(onlypage,text='File: '+pdbfilename).grid(row=0,column=0,sticky='w')
Tkinter.Label(onlypage,text='Atoms: '+selatom).grid(row=1,column=0,sticky='w')
Tkinter.Label(onlypage,text='Analysis: '+noma1).grid(row=2,column=0,sticky='w')
path=os.path.join(os.path.expanduser('~'),'.noma/')
fin = open(path+'savefile.txt','r')
global savedfile
savedfile=fin.readlines()
fin.close()
i=0
a=len(savedfile)
while i<a:
savedfile[i]=savedfile[i][:-1]
i+=1
if gamcut=='0':
Tkinter.Label(onlypage,text='Gamma: r^'+savedfile[91]).grid(row=3,column=0,sticky='w')
Tkinter.Label(onlypage,text='Cutoff: '+cut1).grid(row=4,column=0,sticky='w')
elif gamcut=='1':
Tkinter.Label(onlypage,text='Gamma: '+gam2).grid(row=3,column=0,sticky='w')
Tkinter.Label(onlypage,text='Cutoff: '+cut2).grid(row=4,column=0,sticky='w')
find = 0 #
while find < len(pdbfilename): #
if pdbfilename[-(find+1):-find] == '/': #
bgn = len(pdbfilename)-find #
break #
else: # helps in the
find +=1 # saving of files
try: #
float(bgn) #
except (NameError): #
bgn = 0 #
find = 0 #
while bgn+find<len(pdbfilename): #
if pdbfilename[bgn+find:bgn+find+1] == '.': #
end = len(pdbfilename)-(bgn+find) #
break #
else: #
find +=1 #
try: #
name = pdbfilename[bgn:-end] #
except (NameError): #
name = pdbfilename[bgn:len(pdbfilename)] # name of the file
bgn = pdbfilename[:bgn] # path for file
mytimeis = time.asctime(time.localtime(time.time()))
start = time.time()
try:
p38 = prody.parsePDB(pdbfilename,model=int(modelnumber))
except:
import tkMessageBox
tkMessageBox.askokcancel("File Error","""This is not the correct path or name. Try entering /some/path/nameoffile.pdb
If you need help finding the path, open a new terminal and enter:
find -name 'filename.pdb' use the output as the pdb input
If this doesn't work, make sure the file is in PDB format.""")
p38 = prody.parsePDB(pdbfilename)
print 'Submitted: '+pdbfilename+' at '+mytimeis
Tkinter.Label(onlypage,text='Submitted at: '+mytimeis).grid(row=5,column=0,sticky='w')
root.update()
if selatom == "C-alpha" and noma1 == "Gaussian Normal Mode":
folder = cagnm+'/'
pro = p38.select('protein and name CA') # selects only carbon alpahs
elif selatom == "C-alpha" and noma1 == "Anisotropic Normal Mode":
folder = caanm+'/'
pro = p38.select('protein and name CA')
elif selatom == "Heavy" and noma1 == "Gaussian Normal Mode":
folder = nohgnm+'/'
pro = p38.select('protein and not name "[1-9]?H.*"') # gets rid of all Hydrogens
elif selatom == "Heavy" and noma1 == "Anisotropic Normal Mode":
folder = nohanm+'/'
pro = p38.select('protein and not name "[1-9]?H.*"')
elif selatom == "All" and noma1 == "Gaussian Normal Mode":
folder = allgnm+'/'
pro = p38.select('protein')
elif selatom == "All" and noma1 == "Anisotropic Normal Mode":
folder = allanm+'/'
pro = p38.select('protein')
elif selatom == "Backbone" and noma1 == "Gaussian Normal Mode":
folder = bbgnm+'/'
pro = p38.select('protein and name CA C O N H') # selects backbone
elif selatom == "Backbone" and noma1 == "Anisotropic Normal Mode":
folder = bbanm+'/'
pro = p38.select('protein and name CA C O N H') # selects backbone
elif selatom == "Sidechain" and noma1 == "Gaussian Normal Mode":
folder = scgnm+'/'
pro = p38.select('protein and not name CA C O N H') # selects sidechain
elif selatom == "Sidechain" and noma1 == "Anisotropic Normal Mode":
folder = scanm+'/'
pro = p38.select('protein and not name CA C O N H') # selects sidechain
try: #
open(bgn+folder) # creates the folders
except (IOError): # where the files will
try: # be saved only if they
os.makedirs(bgn+folder) # are not there
except (OSError): #
mer = 0 #
if noma1 == "Gaussian Normal Mode":
print 'Building the Kirchhoff matrix'
Tkinter.Label(onlypage,text='Building Kirchhoff').grid(row=6,column=0,sticky='w')
root.update()
anm = prody.GNM(name)###
if gamcut=='0':
anm.buildKirchhoff(pro,cutoff=float(cut1),gamma=gammaDistanceDependent)###
anm.setKirchhoff(anm.getKirchhoff())
elif gamcut=='1':
anm.buildKirchhoff(pro,cutoff=float(cut2),gamma=float(gam2))###
brat = 2
elif noma1 == "Anisotropic Normal Mode":
print 'Building the Hessian matrix'
Tkinter.Label(onlypage,text='Building Hessian').grid(row=6,column=0,sticky='w')
root.update()
anm = prody.ANM(name)###
if gamcut=='0':
anm.buildHessian(pro,cutoff=float(cut1),gamma=gammaDistanceDependent)###
anm.setHessian(anm.getHessian())###
elif gamcut=='1':
anm.buildHessian(pro,cutoff=float(cut2),gamma=float(gam2))###
brat = 7
print 'Calculating modes'
Tkinter.Label(onlypage,text='Calculating modes').grid(row=7,column=0,sticky='w')
root.update()
anm.calcModes(int(nummodes),zeros = True)###
numatom=anm.numAtoms()###
eigval=anm.getEigvals()###
atomname=pro.getNames()###
if smodel==1:
if brat==2:
modelfilename=bgn+folder+name+modelnewname+'.gnm.npz'
elif brat==7:
modelfilename=bgn+folder+name+modelnewname+'.anm.npz'
print 'Saving Model'
Tkinter.Label(onlypage,text='Saving Model').grid(row=8,column=0,sticky='w')
root.update()
try:
prody.saveModel(anm,bgn+folder+name+modelnewname,True)###
except:
print 'Matrix not saved due to size'
Tkinter.Label(onlypage,text='Matrix not saved').grid(row=8,column=0,sticky='w')
root.update()
prody.saveModel(anm,bgn+folder+name+modelnewname)###
if snmd==1:
print 'Saving NMD'
Tkinter.Label(onlypage,text='Saving NMD').grid(row=9,column=0,sticky='w')
root.update()
try: #
os.makedirs(bgn+folder+nmdfolder+'/') #
except (OSError): #
mer = 0 #
prody.writeNMD(bgn+folder+nmdfolder+'/'+name+nmdnewname+'.nmd',anm[:len(eigval)],pro)### # this can be viewed in VMD
if smodes==1:
print 'Saving Modes'
Tkinter.Label(onlypage,text='Saving Modes').grid(row=10,column=0,sticky='w')
root.update()
try: #
os.makedirs(bgn+folder+modesfolder+'/') #
except (OSError): #
mer = 0 #
modefile = bgn+folder+modesfolder+'/'+name+modesnewname+'.'+modesendname
fout = open(modefile,'w')
mer = 0
while mer< len(eigval):
slowest_mode = anm[mer]###
r = slowest_mode.getEigvec()###
p = slowest_mode.getEigval()###
tq = 0
tt = 0
ttt = 1
tttt = 2
fout.write('MODE {0:3d} {1:15e}'.format(mer+1,p))
fout.write("""
-------------------------------------------------
""")
if noma1 == "Gaussian Normal Mode":
while tq < numatom:
fout.write("""{0:4s}{1:15e}
""".format(atomname[tq],r[tq]))
tq +=1
elif noma1 == "Anisotropic Normal Mode":
while tt < numatom*3:
fout.write("""{0:4s}{1:15e}{2:15e}{3:15e}
""".format(atomname[tq],r[tt],r[ttt],r[tttt]))
tq+=1
tt +=3
ttt+=3
tttt+=3
mer +=1
fout.close()
if showresults=='1':
os.system('/usr/bin/gnome-open '+modefile)
if scollec==1:
print 'Saving collectivity'
Tkinter.Label(onlypage,text='Saving collectivity').grid(row=11,column=0,sticky='w')
root.update()
try: #
os.makedirs(bgn+folder+collectivityfolder+'/') #
except (OSError): #
mer = 0 #
mer = 0
xx = [0]*(numatom) # sets the array to zero and other initial conditions
i = 0
aa = 0
no = 0
var3 = 0
sss = [0]*(len(eigval))
while mer< len(eigval):
slowest_mode = anm[mer]###
r = slowest_mode.getEigvec()###
p = slowest_mode.getEigval()###
a = 0
tt = 0
ttt = 1
tttt = 2
while a < numatom:
atom = atomname[a]
mass = 0
while mass < 2:
if atom[mass] == "N": # all nitrogen
m = 14.0067
break
elif atom[mass] == 'H': # all hydrogen
m = 1.00794
break
elif atom[mass] == "C" : # all carbon
m = 12.0107
break
elif atom[mass] == "O" : # all oxygen
m = 15.9994
break
elif atom[mass] == 'S': # all sulfur
m = 32.065
break
elif atom[mass] == 'P' : # all phosphorus
m = 30.973762
break
else:
if mass == 0:
mass +=1
try:
atom[mass]
except (IndexError):
m = 1
if no == 0:
print 'Enter atom '+atom+' in to the system. Its mass was set to 1 in this simulation.'
no +=1
break
else:
m = 1
if no == 0:
print 'Enter atom '+atom+' in to the system. Its mass was set to 1 in this simulation'
no +=1
break
if len(r)/numatom == 3:
xx[i] = (r[tt]**2 + r[ttt]**2 + r[tttt]**2)/m
i +=1
tt +=3
ttt+=3
tttt+=3
else:
xx[i] = (r[tt]**2)/m
i +=1
tt +=1
a +=1
var3 = 0
j = 0
loop = 1
while loop == 1:
if sum(xx) == 0: # need this because you can't divide by 0
loop = 0
elif j <(numatom):
var1 = xx[j]/sum(xx)
if var1 == 0:
var2 = 0
elif var1 != 0:
from math import log # this means natural log
var2 = var1* log(var1)
var3 += var2
j +=1
else:
from math import exp
k = exp(-var3)/numatom
sss[aa] = k, aa+1
aa +=1
mer +=1
loop = 0
i = 0
xx = [0]*(numatom) # goes through all this until the big loop is done
a = 0
k=[0]*(len(eigval))
while a < len(eigval):
k[a]=prody.calcCollectivity(anm[a]),a+1
a +=1
collectivefile = bgn+folder+collectivityfolder+'/'+name+collectivitynewname+'.'+collectivityendname
fout = open(collectivefile,'w')
if massnomass=='0':
fout.write('MODE COLLECTIVITY(mass)')
fout.write("""
---------------------------
""")
for h in sorted(sss,reverse=True):
fout.write(str(h)[-3:-1]+' '+str(h)[1:19]+"""
""")
fout.write("""
MODE COLLECTIVITY(without mass)""")
fout.write("""
---------------------------
""")
for hh in sorted(k,reverse=True):
fout.write(str(hh)[-3:-1]+' '+str(hh)[1:19]+"""
""")
elif massnomass=='1':
fout.write('MODE COLLECTIVITY(without mass)')
fout.write("""
---------------------------
""")
for hh in sorted(k,reverse=True):
fout.write(str(hh)[-3:-1]+' '+str(hh)[1:19]+"""
""")
fout.write("""
MODE COLLECTIVITY(mass)""")
fout.write("""
---------------------------
""")
for h in sorted(sss,reverse=True):
fout.write(str(h)[-3:-1]+' '+str(h)[1:19]+"""
""")
fout.close()
if showresults=='1':
os.system('/usr/bin/gnome-open '+collectivefile)
fin = open(collectivefile,'r')
lst = fin.readlines()
hi0 = 2
looop = 1
prut=0
secoll=0
thicoll=0
while looop == 1:
fine = lst[hi0]
if int(fine[0:2]) >= brat:
if prut==0:
prut=fine[0:2]
elif secoll==0:
secoll=fine[0:2]
elif thicoll==0:
thicoll=fine[0:2]
else:
foucoll=fine[0:2]
looop = 0
else:
hi0 +=1
mostcollective= "Mode "+prut+" is the most collective."
Tkinter.Label(onlypage,text='Mode '+prut+' is the most collective').grid(row=12,column=0,sticky='w')
root.update()
print mostcollective
fin.close()
if sample1 == 1:
print 'Saving sample file'
Tkinter.Label(onlypage,text='Saving sample file').grid(row=13,column=0,sticky='w')
root.update()
a = modeens+' '
b = [0]*(len(a)+1)
i = 0
j = 0
b1 = 0
while i < len(a):
if a[i:i+1] ==' ' or a[i:i+1]==',':
try:
b[b1]=int(a[j:i])-1
except:
if '1c' in a[j:i]:
b[b1]=int(prut)-1
elif '2c' in a[j:i]:
b[b1]=int(prut)-1
b1 +=1
b[b1]=int(secoll)-1
elif '3c' in a[j:i]:
b[b1]=int(prut)-1
b1 +=1
b[b1]=int(secoll)-1
b1 +=1
b[b1]=int(thicoll)-1
elif '4c' in a[j:i]:
b[b1]=int(prut)-1
b1 +=1
b[b1]=int(secoll)-1
b1 +=1
b[b1]=int(thicoll)-1
b1+=1
b[b1]=int(foucoll)-1
j = i+1
i +=1
b1 +=1
else:
i +=1
del b[b1:]
ensemble = prody.sampleModes(anm[b],pro, n_confs=int(confens), rmsd =float(rmsdens))
p38ens=pro.copy()
p38ens.delCoordset(0)
p38ens.addCoordset(ensemble.getCoordsets())
prody.writePDB(bgn+folder+name+samplenewname+'.pdb',p38ens)
if traverse1 ==1:
print 'Saving traverse file'
Tkinter.Label(onlypage,text='Saving traverse file').grid(row=14,column=0,sticky='w')
root.update()
if modetra=='c':
modefortra=int(prut)-1
else:
modefortra=int(modetra)-1
trajectory=prody.traverseMode(anm[modefortra],pro,n_steps=int(steptra),rmsd=float(rmsdtra))
prody.calcRMSD(trajectory).round(2)
p38traj=pro.copy()
p38traj.delCoordset(0)
p38traj.addCoordset(trajectory.getCoordsets())
prody.writePDB(bgn+folder+name+'_mode'+str(modefortra+1)+traversenewname+'.pdb',p38traj)
if crosscorr==1:
print 'Saving cross correlation'
Tkinter.Label(onlypage,text='Saving cross-correlation').grid(row=15,column=0,sticky='w')
root.update()
try: #
os.makedirs(bgn+folder+corrfolder+'/') #
except (OSError): #
mer = 0
i=int(compmode01)
while i <= int(compmode02):
x=i-1
correlationdataname=bgn+folder+corrfolder+'/'+name+corrname+'_mode'+str(x+1)+'.'+corrend
prody.writeArray(correlationdataname,prody.calcCrossCorr(anm[x]),'%.18e')
print correlationdataname
i+=1
##
if sqflucts==1:
print 'Saving square fluctuation'
Tkinter.Label(onlypage,text='Saving square fluctuation').grid(row=16,column=0,sticky='w')
root.update()
try: #
os.makedirs(bgn+folder+sqfluctsfolder+'/') #
except (OSError): #
mer = 0
i=int(compmode01)
while i < int(compmode02):
yelp = i-1
sqfluctdataname = bgn+folder+sqfluctsfolder+'/'+name+sqfluctsname+'_mode'+str(yelp+1)+'.'+sqfluctsend
fout = open(sqfluctdataname,'w')
if separatevar1=='0':
a = 0
while a < numatom:
fout.write(str(a))
fout.write(""" """)
fout.write(str(prody.calcSqFlucts(anm[yelp])[a]))
fout.write("""
""")
a +=1
elif separatevar1=='1':
a=0
while a <numatom:
firstresnum=int(p38.getResnums()[0:1][0])
origiresnum=int(p38.getResnums()[0:1][0])
while firstresnum<(int(numatom*1.0/p38.numChains())+origiresnum):
fout.write(str(firstresnum))
fout.write('\t')
fout.write(str(prody.calcSqFlucts(anm[yelp])[a]))
fout.write('\n')
a+=1
firstresnum+=1
fout.write('&\n')
fout.close()
print sqfluctdataname
i+=1
if temfac==1:
print 'Saving temperature factors'
Tkinter.Label(onlypage,text='Saving temperature factors').grid(row=17,column=0,sticky='w')
root.update()
try: #
os.makedirs(bgn+folder+temfacfolder+'/') #
except (OSError): #
mer = 0
fin=open(pdbfilename,'r')
d = [None]*len(atomname)
e = 0
for line in fin:
pair = line.split()
if 'ATOM ' in line and e < len(atomname):
if str(pair[2]) == str(atomname[e]):
d[e]=str(pair[1])
e+=1
else:
e+=0
else:
continue
fin.close()
sqf = prody.calcSqFlucts(anm)
x = sqf/((sqf**2).sum()**.5)
y = prody.calcTempFactors(anm,pro)
a = 0
tempfactorsdataname =bgn+folder+temfacfolder+'/'+name+temfacname+'.'+temfacend
fout=open(tempfactorsdataname,'w')
fout.write("""Atom Residue TempFactor TempFactor with exp beta
""")
while a < numatom:
fout.write("""{0:4s} {1:4d} {2:15f} {3:15f}
""".format(d[a],a+1,x[a],y[a]))
a +=1
fout.close()
print tempfactorsdataname
if fracovar==1:
try:
import matplotlib.pyplot as plt
print 'Saving Fraction of Variance'
Tkinter.Label(onlypage,text='Saving Fraction of Variance').grid(row=18,column=0,sticky='w')
root.update()
try: #
os.makedirs(bgn+folder+modesfolder+'/') #
except (OSError): #
mer = 0 #
plt.figure(figsize = (5,4))
prody.showFractVars(anm)
prody.showCumulFractVars(anm)
fracvardataname =bgn+folder+modesfolder+'/'+name+fraconame+'.'+fracoend
plt.savefig(fracvardataname)
print fracvardataname
if showresults=='1':
os.system('/usr/bin/gnome-open '+fracvardataname)
except:
print 'Error: Fraction of Variance'
Tkinter.Label(onlypage,text='Error: Fraction of Variance').grid(row=18,column=0,sticky='w')
root.update()
mer=0
if ovlap==1 or ovlaptab==1:
try:
import matplotlib.pyplot as plt
print 'Saving Overlap'
Tkinter.Label(onlypage,text='Saving Overlap').grid(row=19,column=0,sticky='w')
root.update()
Tkinter.Label(onlypage,text='Comparison: '+comppdbfilename).grid(row=20,column=0,sticky='w')
##
find = 0
while find < len(comppdbfilename):
if comppdbfilename[-(find+1):-find] == '/':
bgn1 = len(comppdbfilename)-find
break
else:
find +=1
try:
float(bgn1)
except (NameError):
bgn1 = 0
find = 0
while bgn1+find<len(comppdbfilename):
if comppdbfilename[bgn1+find:bgn1+find+1] == '.':
end1 = len(comppdbfilename)-(bgn1+find)
break
else:
find +=1
try:
name1 = comppdbfilename[bgn1:-end1]
except (NameError):
name1 = comppdbfilename[bgn1:len(comppdbfilename)]
bgn1 = comppdbfilename[:bgn1]
p381 = prody.parsePDB(comppdbfilename,model=int(modelnumber))
if selatom == "C-alpha" and noma1 == "Gaussian Normal Mode":
pro1 = p381.select('protein and name CA')
elif selatom == "C-alpha" and noma1 == "Anisotropic Normal Mode":
pro1 = p381.select('protein and name CA')
elif selatom == "Heavy" and noma1 == "Gaussian Normal Mode":
pro1 = p381.select('protein and not name "[1-9]?H.*"')
elif selatom == "Heavy" and noma1 == "Anisotropic Normal Mode":
pro1 = p381.select('protein and not name "[1-9]?H.*"')
elif selatom == "All" and noma1 == "Gaussian Normal Mode":
pro1 = p381.select('protein')
elif selatom == "All" and noma1 == "Anisotropic Normal Mode":
pro1 = p381.select('protein')
elif selatom == "Backbone" and noma1 == "Gaussian Normal Mode":
pro1 = p381.select('protein and name CA C O N H')
elif selatom == "Backbone" and noma1 == "Anisotropic Normal Mode":
pro1 = p381.select('protein and name CA C O N H')
elif selatom == "Sidechain" and noma1 == "Gaussian Normal Mode":
pro1 = p381.select('protein and not name CA C O N H')
elif selatom == "Sidechain" and noma1 == "Anisotropic Normal Mode":
pro1 = p381.select('protein and not name CA C O N H')
if noma1 == "Gaussian Normal Mode":
print 'Building the Kirchhoff matrix'
Tkinter.Label(onlypage,text='Building Kirchhoff').grid(row=21,column=0,sticky='w')
root.update()
anm1 = prody.GNM(name1)
if gamcut=='0':
anm1.buildKirchhoff(pro1,cutoff=float(cut1),gamma=gammaDistanceDependent)
anm1.setKirchhoff(anm1.getKirchhoff())
elif gamcut=='1':
anm1.buildKirchhoff(pro1,cutoff=float(cut2),gamma=float(gam2))
brat = 2
elif noma1 == "Anisotropic Normal Mode":
print 'Building the Hessian matrix'
Tkinter.Label(onlypage,text='Building Hessian').grid(row=21,column=0,sticky='w')
root.update()
anm1 = prody.ANM(name1)
if gamcut=='0':
anm1.buildHessian(pro1,cutoff=float(cut1),gamma=gammaDistanceDependent)
anm1.setHessian(anm1.getHessian())
elif gamcut=='1':
anm1.buildHessian(pro1,cutoff=float(cut2),gamma=float(gam2))
brat = 7
print 'Calculating modes'
Tkinter.Label(onlypage,text='Calculating modes').grid(row=22,column=0,sticky='w')
root.update()
anm1.calcModes(int(nummodes),zeros = True)
##
try:
os.makedirs(bgn+folder+ovlapfold+'/')
except (OSError):
mer = 0
if ovlap==1:
i=int(compmode01)
while i < int(compmode02):
a = i-1
plt.figure(figsize=(5,4))
prody.showCumulOverlap(anm[a],anm1)
prody.showOverlap(anm[a],anm1)
plt.title('Overlap with Mode '+str(a+1)+' from '+name)
plt.xlabel(name1+' mode index')
overlapname = bgn+folder+ovlapfold+'/'+name+'_'+name1+ovlapname+'_mode'+str(a+1)+'.'+ovlapend
plt.savefig(overlapname)
print overlapname
i+=1
if ovlaptab==1:
plt.figure(figsize=(5,4))
prody.showOverlapTable(anm1,anm)
plt.xlim(int(compmode01)-1,int(compmode02))
plt.ylim(int(compmode01)-1,int(compmode02))
plt.title(name1+' vs '+name+' Overlap')
plt.ylabel(name1)
plt.xlabel(name)
overlapname = bgn+folder+ovlapfold+'/'+name+'_'+name1+ovlaptabname+'.'+ovlaptabend
plt.savefig(overlapname)
print overlapname
except:
mer=0
root.destroy()
mynewtimeis = float(time.time()-start)
if mynewtimeis <= 60.00:
timeittook= "The calculations took %.2f s."%(mynewtimeis)
elif mynewtimeis > 60.00 and mynewtimeis <= 3600.00:
timeittook= "The calculations took %.2f min."%((mynewtimeis/60.00))
else:
timeittook= "The calculations took %.2f hrs."%((mynewtimeis/3600.00))
print timeittook
if smodel==1 and scollec==1:
return (timeittook,modelfilename,str(int(prut)))
elif scollec==1:
return (timeittook,'nofile',str(int(prut)))
elif smodel==1:
return (timeittook,modelfilename,'nocoll')
else:
return (timeittook,'nofile','nocoll')
def prody_gnm(pdb, **kwargs):
"""Perform GNM calculations for *pdb*.
"""
for key in DEFAULTS:
if not key in kwargs:
kwargs[key] = DEFAULTS[key]
from os.path import isdir, splitext, join
outdir = kwargs.get("outdir")
if not isdir(outdir):
raise IOError("{0} is not a valid path".format(repr(outdir)))
import numpy as np
import prody
LOGGER = prody.LOGGER
selstr = kwargs.get("select")
prefix = kwargs.get("prefix")
cutoff = kwargs.get("cutoff")
gamma = kwargs.get("gamma")
nmodes = kwargs.get("nmodes")
selstr = kwargs.get("select")
model = kwargs.get("model")
pdb = prody.parsePDB(pdb, model=model)
if prefix == "_gnm":
prefix = pdb.getTitle() + "_gnm"
select = pdb.select(selstr)
if select is None:
raise ValueError("selection {0} do not match any atoms".format(repr(selstr)))
LOGGER.info("{0} atoms will be used for GNM calculations.".format(len(select)))
gnm = prody.GNM(pdb.getTitle())
gnm.buildKirchhoff(select, cutoff, gamma)
gnm.calcModes(nmodes)
LOGGER.info("Writing numerical output.")
if kwargs.get("outnpz"):
prody.saveModel(gnm, join(outdir, prefix))
prody.writeNMD(join(outdir, prefix + ".nmd"), gnm, select)
extend = kwargs.get("extend")
if extend:
if extend == "all":
extended = prody.extendModel(gnm, select, pdb)
else:
extended = prody.extendModel(gnm, select, select | pdb.bb)
prody.writeNMD(join(outdir, prefix + "_extended_" + extend + ".nmd"), *extended)
outall = kwargs.get("outall")
delim = kwargs.get("numdelim")
ext = kwargs.get("numext")
format = kwargs.get("numformat")
if outall or kwargs.get("outeig"):
prody.writeArray(join(outdir, prefix + "_evectors" + ext), gnm.getArray(), delimiter=delim, format=format)
prody.writeArray(join(outdir, prefix + "_evalues" + ext), gnm.getEigvals(), delimiter=delim, format=format)
if outall or kwargs.get("outbeta"):
from prody.utilities import openFile
fout = openFile(prefix + "_beta.txt", "w", folder=outdir)
fout.write("{0[0]:1s} {0[1]:4s} {0[2]:4s} {0[3]:5s} {0[4]:5s}\n".format(["C", "RES", "####", "Exp.", "The."]))
for data in zip(
select.getChids(),
select.getResnames(),
select.getResnums(),
select.getBetas(),
prody.calcTempFactors(gnm, select),
):
fout.write("{0[0]:1s} {0[1]:4s} {0[2]:4d} {0[3]:5.2f} {0[4]:5.2f}\n".format(data))
fout.close()
if outall or kwargs.get("outcov"):
prody.writeArray(
join(outdir, prefix + "_covariance" + ext), gnm.getCovariance(), delimiter=delim, format=format
)
if outall or kwargs.get("outcc") or kwargs.get("outhm"):
cc = prody.calcCrossCorr(gnm)
if outall or kwargs.get("outcc"):
prody.writeArray(join(outdir, prefix + "_cross-correlations" + ext), cc, delimiter=delim, format=format)
if outall or kwargs.get("outhm"):
prody.writeHeatmap(
join(outdir, prefix + "_cross-correlations.hm"),
cc,
resnum=select.getResnums(),
xlabel="Residue",
ylabel="Residue",
title=gnm.getTitle() + " cross-correlations",
)
if outall or kwargs.get("kirchhoff"):
prody.writeArray(join(outdir, prefix + "_kirchhoff" + ext), gnm.getKirchhoff(), delimiter=delim, format=format)
if outall or kwargs.get("outsf"):
prody.writeArray(
join(outdir, prefix + "_sqfluct" + ext), prody.calcSqFlucts(gnm), delimiter=delim, format=format
)
figall = kwargs.get("figall")
cc = kwargs.get("figcc")
sf = kwargs.get("figsf")
bf = kwargs.get("figbeta")
cm = kwargs.get("figcmap")
modes = kwargs.get("figmode")
if figall or cc or sf or bf or cm or modes:
try:
import matplotlib.pyplot as plt
except ImportError:
LOGGER.warning("Matplotlib could not be imported. " "Figures are not saved.")
else:
prody.SETTINGS["auto_show"] = False
LOGGER.info("Saving graphical output.")
format = kwargs.get("figformat")
width = kwargs.get("figwidth")
height = kwargs.get("figheight")
dpi = kwargs.get("figdpi")
format = format.lower()
if figall or cc:
plt.figure(figsize=(width, height))
prody.showCrossCorr(gnm)
plt.savefig(join(outdir, prefix + "_cc." + format), dpi=dpi, format=format)
plt.close("all")
if figall or cm:
plt.figure(figsize=(width, height))
prody.showContactMap(gnm)
plt.savefig(join(outdir, prefix + "_cm." + format), dpi=dpi, format=format)
plt.close("all")
if figall or sf:
plt.figure(figsize=(width, height))
prody.showSqFlucts(gnm)
plt.savefig(join(outdir, prefix + "_sf." + format), dpi=dpi, format=format)
plt.close("all")
if figall or bf:
plt.figure(figsize=(width, height))
bexp = select.getBetas()
bcal = prody.calcTempFactors(gnm, select)
plt.plot(bexp, label="Experimental")
plt.plot(bcal, label=("Theoretical (corr coef = {0:.2f})".format(np.corrcoef(bcal, bexp)[0, 1])))
plt.legend(prop={"size": 10})
plt.xlabel("Node index")
plt.ylabel("Experimental B-factors")
plt.title(pdb.getTitle() + " B-factors")
plt.savefig(join(outdir, prefix + "_bf." + format), dpi=dpi, format=format)
plt.close("all")
if modes:
indices = []
items = modes.split()
items = sum([item.split(",") for item in items], [])
for item in items:
try:
item = item.split("-")
if len(item) == 1:
indices.append(int(item[0]) - 1)
elif len(item) == 2:
indices.extend(range(int(item[0]) - 1, int(item[1])))
except:
pass
for index in indices:
try:
mode = gnm[index]
except:
pass
else:
plt.figure(figsize=(width, height))
prody.showMode(mode)
plt.grid()
plt.savefig(
join(outdir, prefix + "_mode_" + str(mode.getIndex() + 1) + "." + format),
dpi=dpi,
format=format,
)
plt.close("all")
def get_gnm_fluctuations(protein: pd.AtomGroup, n_modes: int = 50):
"""
Get atom fluctuations using a Gaussian network model with n_modes modes.
"""
protein_gnm, _ = pd.calcGNM(protein, n_modes=n_modes, selstr="all")
return pd.calcSqFlucts(protein_gnm)
def get_anm_fluctuations(protein: pd.AtomGroup, n_modes: int = 50):
"""
Get atom fluctuations using an Anisotropic network model with n_modes modes.
"""
protein_anm, _ = pd.calcANM(protein, n_modes=n_modes, selstr="all")
return pd.calcSqFlucts(protein_anm)
def prody_gnm(pdb, **kwargs):
"""Perform GNM calculations for *pdb*.
"""
for key in DEFAULTS:
if not key in kwargs:
kwargs[key] = DEFAULTS[key]
from os.path import isdir, splitext, join
outdir = kwargs.get('outdir')
if not isdir(outdir):
raise IOError('{0} is not a valid path'.format(repr(outdir)))
import numpy as np
import prody
LOGGER = prody.LOGGER
selstr = kwargs.get('select')
prefix = kwargs.get('prefix')
cutoff = kwargs.get('cutoff')
gamma = kwargs.get('gamma')
nmodes = kwargs.get('nmodes')
selstr = kwargs.get('select')
model = kwargs.get('model')
altloc = kwargs.get('altloc')
zeros = kwargs.get('zeros')
pdb = prody.parsePDB(pdb, model=model, altloc=altloc)
if prefix == '_gnm':
prefix = pdb.getTitle() + '_gnm'
select = pdb.select(selstr)
if select is None:
raise ValueError('selection {0} do not match any atoms'.format(
repr(selstr)))
LOGGER.info('{0} atoms will be used for GNM calculations.'.format(
len(select)))
gnm = prody.GNM(pdb.getTitle())
nproc = kwargs.get('nproc')
if nproc:
try:
from threadpoolctl import threadpool_limits
except ImportError:
raise ImportError(
'Please install threadpoolctl to control threads')
with threadpool_limits(limits=nproc, user_api="blas"):
gnm.buildKirchhoff(select, cutoff, gamma)
gnm.calcModes(nmodes, zeros=zeros)
else:
gnm.buildKirchhoff(select, cutoff, gamma)
gnm.calcModes(nmodes, zeros=zeros)
LOGGER.info('Writing numerical output.')
if kwargs.get('outnpz'):
prody.saveModel(gnm, join(outdir, prefix))
if kwargs.get('outscipion'):
prody.writeScipionModes(outdir, gnm)
prody.writeNMD(join(outdir, prefix + '.nmd'), gnm, select)
extend = kwargs.get('extend')
if extend:
if extend == 'all':
extended = prody.extendModel(gnm, select, pdb)
else:
extended = prody.extendModel(gnm, select, select | pdb.bb)
prody.writeNMD(join(outdir, prefix + '_extended_' + extend + '.nmd'),
*extended)
outall = kwargs.get('outall')
delim = kwargs.get('numdelim')
ext = kwargs.get('numext')
format = kwargs.get('numformat')
if outall or kwargs.get('outeig'):
prody.writeArray(join(outdir, prefix + '_evectors' + ext),
gnm.getArray(),
delimiter=delim,
format=format)
prody.writeArray(join(outdir, prefix + '_evalues' + ext),
gnm.getEigvals(),
delimiter=delim,
format=format)
if outall or kwargs.get('outbeta'):
from prody.utilities import openFile
fout = openFile(prefix + '_beta' + ext, 'w', folder=outdir)
fout.write(
'{0[0]:1s} {0[1]:4s} {0[2]:4s} {0[3]:5s} {0[4]:5s}\n'.format(
['C', 'RES', '####', 'Exp.', 'The.']))
for data in zip(select.getChids(), select.getResnames(),
select.getResnums(), select.getBetas(),
prody.calcTempFactors(gnm, select)):
fout.write(
'{0[0]:1s} {0[1]:4s} {0[2]:4d} {0[3]:5.2f} {0[4]:5.2f}\n'.
format(data))
fout.close()
if outall or kwargs.get('outcov'):
prody.writeArray(join(outdir, prefix + '_covariance' + ext),
gnm.getCovariance(),
delimiter=delim,
format=format)
if outall or kwargs.get('outcc') or kwargs.get('outhm'):
cc = prody.calcCrossCorr(gnm)
if outall or kwargs.get('outcc'):
prody.writeArray(join(outdir,
prefix + '_cross-correlations' + ext),
cc,
delimiter=delim,
format=format)
if outall or kwargs.get('outhm'):
prody.writeHeatmap(join(outdir, prefix + '_cross-correlations.hm'),
cc,
resnum=select.getResnums(),
xlabel='Residue',
ylabel='Residue',
title=gnm.getTitle() + ' cross-correlations')
if outall or kwargs.get('kirchhoff'):
prody.writeArray(join(outdir, prefix + '_kirchhoff' + ext),
gnm.getKirchhoff(),
delimiter=delim,
format=format)
if outall or kwargs.get('outsf'):
prody.writeArray(join(outdir, prefix + '_sqfluct' + ext),
prody.calcSqFlucts(gnm),
delimiter=delim,
format=format)
figall = kwargs.get('figall')
cc = kwargs.get('figcc')
sf = kwargs.get('figsf')
bf = kwargs.get('figbeta')
cm = kwargs.get('figcmap')
modes = kwargs.get('figmode')
if figall or cc or sf or bf or cm or modes:
try:
import matplotlib.pyplot as plt
except ImportError:
LOGGER.warning('Matplotlib could not be imported. '
'Figures are not saved.')
else:
prody.SETTINGS['auto_show'] = False
LOGGER.info('Saving graphical output.')
format = kwargs.get('figformat')
width = kwargs.get('figwidth')
height = kwargs.get('figheight')
dpi = kwargs.get('figdpi')
format = format.lower()
if figall or cc:
plt.figure(figsize=(width, height))
prody.showCrossCorr(gnm)
plt.savefig(join(outdir, prefix + '_cc.' + format),
dpi=dpi,
format=format)
plt.close('all')
if figall or cm:
plt.figure(figsize=(width, height))
prody.showContactMap(gnm)
plt.savefig(join(outdir, prefix + '_cm.' + format),
dpi=dpi,
format=format)
plt.close('all')
if figall or sf:
plt.figure(figsize=(width, height))
prody.showSqFlucts(gnm)
plt.savefig(join(outdir, prefix + '_sf.' + format),
dpi=dpi,
format=format)
plt.close('all')
if figall or bf:
plt.figure(figsize=(width, height))
bexp = select.getBetas()
bcal = prody.calcTempFactors(gnm, select)
plt.plot(bexp, label='Experimental')
plt.plot(bcal,
label=('Theoretical (corr coef = {0:.2f})'.format(
np.corrcoef(bcal, bexp)[0, 1])))
plt.legend(prop={'size': 10})
plt.xlabel('Node index')
plt.ylabel('Experimental B-factors')
plt.title(pdb.getTitle() + ' B-factors')
plt.savefig(join(outdir, prefix + '_bf.' + format),
dpi=dpi,
format=format)
plt.close('all')
if modes:
indices = []
items = modes.split()
items = sum([item.split(',') for item in items], [])
for item in items:
try:
item = item.split('-')
if len(item) == 1:
indices.append(int(item[0]) - 1)
elif len(item) == 2:
indices.extend(
list(range(int(item[0]) - 1, int(item[1]))))
except:
pass
for index in indices:
try:
mode = gnm[index]
except:
pass
else:
plt.figure(figsize=(width, height))
prody.showMode(mode)
plt.grid()
plt.savefig(join(
outdir, prefix + '_mode_' +
str(mode.getIndex() + 1) + '.' + format),
dpi=dpi,
format=format)
plt.close('all')
def calcANMfeatures(self, chain='all', env='chain',
ANM_PRS=True, stiffness=True, MBS=False):
"""Computes ANM-based features.
:arg chain: chain identifier
:type chain: str
:arg env: environment model, i.e. ``'chain'``, ``'reduced'`` or
``'sliced'``
:type env: str
:arg ANM_PRS: whether or not to compute features based on Perturbation
Response Scanning analysis
:type ANM_PRS: bool
:arg stiffness: whether or not to compute stiffness with MechStiff
:type stiffness: bool
:arg MBS: whether or not to compute Mechanical Bridging Score
:type MBS: bool
"""
assert env in ['chain', 'reduced', 'sliced']
for k in ANM_PRS, stiffness, MBS:
assert type(k) is bool
# list of features to be computed
features = ['ANM_MSF-'+env]
if ANM_PRS:
features += ['ANM_effectiveness-'+env, 'ANM_sensitivity-'+env]
if MBS:
features += ['MBS-'+env]
if stiffness:
features += ['stiffness-'+env]
# compute features (if not precomputed)
if chain == 'all':
chain_list = self.chids
else:
chain_list = [chain, ]
for chID in chain_list:
d = self.feats[chID]
if all([f in d for f in features]):
continue
try:
anm = self.calcANM(chID, env=env)
except Exception as e:
if (isinstance(e, MemoryError)):
msg = 'MemoryError'
else:
msg = str(e)
for f in features:
d[f] = msg
LOGGER.warn(msg)
continue
key_msf = 'ANM_MSF-' + env
if key_msf not in d:
try:
d[key_msf] = calcSqFlucts(anm)
except Exception as e:
msg = str(e)
d[key_msf] = msg
LOGGER.warn(msg)
key_eff = 'ANM_effectiveness-' + env
if key_eff in features and key_eff not in d:
key_sns = 'ANM_sensitivity-' + env
try:
prs_mtrx, eff, sns = calcPerturbResponse(anm)
d[key_eff] = eff
d[key_sns] = sns
except Exception as e:
msg = str(e)
d[key_eff] = msg
d[key_sns] = msg
LOGGER.warn(msg)
key_mbs = 'MBS-' + env
if key_mbs in features and key_mbs not in d:
try:
pdb = self.getPDB()
ca = pdb[chID].ca
d[key_mbs] = calcMBS(anm, ca, cutoff=15.)
except Exception as e:
msg = str(e)
d[key_mbs] = msg
LOGGER.warn(msg)
key_stf = 'stiffness-' + env
if key_stf in features and key_stf not in d:
try:
pdb = self.getPDB()
ca = pdb[chID].ca
stiff_mtrx = calcMechStiff(anm, ca)
d[key_stf] = np.mean(stiff_mtrx, axis=0)
except Exception as e:
msg = str(e)
d[key_stf] = msg
LOGGER.warn(msg)
return
def prody_pca(coords, **kwargs):
"""Perform PCA calculations for PDB or DCD format *coords* file.
"""
for key in DEFAULTS:
if not key in kwargs:
kwargs[key] = DEFAULTS[key]
from os.path import isdir, splitext, join
outdir = kwargs.get('outdir')
if not isdir(outdir):
raise IOError('{0} is not a valid path'.format(repr(outdir)))
import prody
LOGGER = prody.LOGGER
prefix = kwargs.get('prefix')
nmodes = kwargs.get('nmodes')
selstr = kwargs.get('select')
quiet = kwargs.pop('quiet', False)
altloc = kwargs.get('altloc')
ext = splitext(coords)[1].lower()
if ext == '.gz':
ext = splitext(coords[:-3])[1].lower()
if ext == '.dcd':
pdb = kwargs.get('psf') or kwargs.get('pdb')
if pdb:
if splitext(pdb)[1].lower() == '.psf':
pdb = prody.parsePSF(pdb)
else:
pdb = prody.parsePDB(pdb, altlocs=altlocs)
dcd = prody.DCDFile(coords)
if prefix == '_pca' or prefix == '_eda':
prefix = dcd.getTitle() + prefix
if len(dcd) < 2:
raise ValueError('DCD file must have multiple frames')
if pdb:
if pdb.numAtoms() == dcd.numAtoms():
select = pdb.select(selstr)
dcd.setAtoms(select)
LOGGER.info('{0} atoms are selected for calculations.'.format(
len(select)))
else:
select = pdb.select(selstr)
if select.numAtoms() != dcd.numAtoms():
raise ValueError('number of selected atoms ({0}) does '
'not match number of atoms in the DCD '
'file ({1})'.format(
select.numAtoms(), dcd.numAtoms()))
if pdb.numCoordsets():
dcd.setCoords(select.getCoords())
else:
select = prody.AtomGroup()
select.setCoords(dcd.getCoords())
pca = prody.PCA(dcd.getTitle())
nproc = kwargs.get('nproc')
if nproc:
try:
from threadpoolctl import threadpool_limits
except ImportError:
raise ImportError(
'Please install threadpoolctl to control threads')
with threadpool_limits(limits=nproc, user_api="blas"):
if len(dcd) > 1000:
pca.buildCovariance(dcd,
aligned=kwargs.get('aligned'),
quiet=quiet)
pca.calcModes(nmodes)
ensemble = dcd
else:
ensemble = dcd[:]
if not kwargs.get('aligned'):
ensemble.iterpose(quiet=quiet)
pca.performSVD(ensemble)
nmodes = pca.numModes()
else:
if len(dcd) > 1000:
pca.buildCovariance(dcd,
aligned=kwargs.get('aligned'),
quiet=quiet)
pca.calcModes(nmodes)
ensemble = dcd
else:
ensemble = dcd[:]
if not kwargs.get('aligned'):
ensemble.iterpose(quiet=quiet)
pca.performSVD(ensemble)
nmodes = pca.numModes()
else:
pdb = prody.parsePDB(coords)
if pdb.numCoordsets() < 2:
raise ValueError('PDB file must contain multiple models')
if prefix == '_pca' or prefix == '_eda':
prefix = pdb.getTitle() + prefix
select = pdb.select(selstr)
LOGGER.info('{0} atoms are selected for calculations.'.format(
len(select)))
if select is None:
raise ValueError('selection {0} do not match any atoms'.format(
repr(selstr)))
LOGGER.info('{0} atoms will be used for PCA calculations.'.format(
len(select)))
ensemble = prody.Ensemble(select)
pca = prody.PCA(pdb.getTitle())
if not kwargs.get('aligned'):
ensemble.iterpose()
nproc = kwargs.get('nproc')
if nproc:
try:
from threadpoolctl import threadpool_limits
except ImportError:
raise ImportError(
'Please install threadpoolctl to control threads')
with threadpool_limits(limits=nproc, user_api="blas"):
pca.performSVD(ensemble)
else:
pca.performSVD(ensemble)
LOGGER.info('Writing numerical output.')
if kwargs.get('outnpz'):
prody.saveModel(pca, join(outdir, prefix))
if kwargs.get('outscipion'):
prody.writeScipionModes(outdir, pca)
prody.writeNMD(join(outdir, prefix + '.nmd'), pca[:nmodes], select)
extend = kwargs.get('extend')
if extend:
if pdb:
if extend == 'all':
extended = prody.extendModel(pca[:nmodes], select, pdb)
else:
extended = prody.extendModel(pca[:nmodes], select,
select | pdb.bb)
prody.writeNMD(
join(outdir, prefix + '_extended_' + extend + '.nmd'),
*extended)
else:
prody.LOGGER.warn('Model could not be extended, provide a PDB or '
'PSF file.')
outall = kwargs.get('outall')
delim = kwargs.get('numdelim')
ext = kwargs.get('numext')
format = kwargs.get('numformat')
if outall or kwargs.get('outeig'):
prody.writeArray(join(outdir, prefix + '_evectors' + ext),
pca.getArray(),
delimiter=delim,
format=format)
prody.writeArray(join(outdir, prefix + '_evalues' + ext),
pca.getEigvals(),
delimiter=delim,
format=format)
if outall or kwargs.get('outcov'):
prody.writeArray(join(outdir, prefix + '_covariance' + ext),
pca.getCovariance(),
delimiter=delim,
format=format)
if outall or kwargs.get('outcc') or kwargs.get('outhm'):
cc = prody.calcCrossCorr(pca)
if outall or kwargs.get('outcc'):
prody.writeArray(join(outdir,
prefix + '_cross-correlations' + ext),
cc,
delimiter=delim,
format=format)
if outall or kwargs.get('outhm'):
resnums = select.getResnums()
hmargs = {} if resnums is None else {'resnums': resnums}
prody.writeHeatmap(join(outdir, prefix + '_cross-correlations.hm'),
cc,
xlabel='Residue',
ylabel='Residue',
title=pca.getTitle() + ' cross-correlations',
**hmargs)
if outall or kwargs.get('outsf'):
prody.writeArray(join(outdir, prefix + '_sqfluct' + ext),
prody.calcSqFlucts(pca),
delimiter=delim,
format=format)
if outall or kwargs.get('outproj'):
prody.writeArray(join(outdir, prefix + '_proj' + ext),
prody.calcProjection(ensemble, pca),
delimiter=delim,
format=format)
figall = kwargs.get('figall')
cc = kwargs.get('figcc')
sf = kwargs.get('figsf')
sp = kwargs.get('figproj')
if figall or cc or sf or sp:
try:
import matplotlib.pyplot as plt
except ImportError:
LOGGER.warning('Matplotlib could not be imported. '
'Figures are not saved.')
else:
prody.SETTINGS['auto_show'] = False
LOGGER.info('Saving graphical output.')
format = kwargs.get('figformat')
width = kwargs.get('figwidth')
height = kwargs.get('figheight')
dpi = kwargs.get('figdpi')
format = format.lower()
if figall or cc:
plt.figure(figsize=(width, height))
prody.showCrossCorr(pca)
plt.savefig(join(outdir, prefix + '_cc.' + format),
dpi=dpi,
format=format)
plt.close('all')
if figall or sf:
plt.figure(figsize=(width, height))
prody.showSqFlucts(pca)
plt.savefig(join(outdir, prefix + '_sf.' + format),
dpi=dpi,
format=format)
plt.close('all')
if figall or sp:
indices = []
for item in sp.split():
try:
if '-' in item:
item = item.split('-')
if len(item) == 2:
indices.append(
list(range(int(item[0]) - 1,
int(item[1]))))
elif ',' in item:
indices.append(
[int(i) - 1 for i in item.split(',')])
else:
indices.append(int(item) - 1)
except:
pass
for index in indices:
plt.figure(figsize=(width, height))
prody.showProjection(ensemble, pca[index])
if isinstance(index, Integral):
index = [index]
index = [str(i + 1) for i in index]
plt.savefig(join(
outdir,
prefix + '_proj_' + '_'.join(index) + '.' + format),
dpi=dpi,
format=format)
plt.close('all')
def prody_anm(pdb, **kwargs):
"""Perform ANM calculations for *pdb*.
"""
for key in DEFAULTS:
if not key in kwargs:
kwargs[key] = DEFAULTS[key]
from os.path import isdir, join
outdir = kwargs.get('outdir')
if not isdir(outdir):
raise IOError('{0} is not a valid path'.format(repr(outdir)))
import numpy as np
import prody
LOGGER = prody.LOGGER
selstr = kwargs.get('select')
prefix = kwargs.get('prefix')
cutoff = kwargs.get('cutoff')
gamma = kwargs.get('gamma')
nmodes = kwargs.get('nmodes')
selstr = kwargs.get('select')
model = kwargs.get('model')
pdb = prody.parsePDB(pdb, model=model)
if prefix == '_anm':
prefix = pdb.getTitle() + '_anm'
select = pdb.select(selstr)
if select is None:
LOGGER.warn('Selection {0} did not match any atoms.'
.format(repr(selstr)))
return
LOGGER.info('{0} atoms will be used for ANM calculations.'
.format(len(select)))
anm = prody.ANM(pdb.getTitle())
anm.buildHessian(select, cutoff, gamma)
anm.calcModes(nmodes)
LOGGER.info('Writing numerical output.')
if kwargs.get('outnpz'):
prody.saveModel(anm, join(outdir, prefix))
prody.writeNMD(join(outdir, prefix + '.nmd'), anm, select)
extend = kwargs.get('extend')
if extend:
if extend == 'all':
extended = prody.extendModel(anm, select, pdb)
else:
extended = prody.extendModel(anm, select, select | pdb.bb)
prody.writeNMD(join(outdir, prefix + '_extended_' +
extend + '.nmd'), *extended)
outall = kwargs.get('outall')
delim = kwargs.get('numdelim')
ext = kwargs.get('numext')
format = kwargs.get('numformat')
if outall or kwargs.get('outeig'):
prody.writeArray(join(outdir, prefix + '_evectors'+ext),
anm.getArray(), delimiter=delim, format=format)
prody.writeArray(join(outdir, prefix + '_evalues'+ext),
anm.getEigvals(), delimiter=delim, format=format)
if outall or kwargs.get('outbeta'):
from prody.utilities import openFile
fout = openFile(prefix + '_beta.txt', 'w', folder=outdir)
fout.write('{0[0]:1s} {0[1]:4s} {0[2]:4s} {0[3]:5s} {0[4]:5s}\n'
.format(['C', 'RES', '####', 'Exp.', 'The.']))
for data in zip(select.getChids(), select.getResnames(),
select.getResnums(), select.getBetas(),
prody.calcTempFactors(anm, select)):
fout.write('{0[0]:1s} {0[1]:4s} {0[2]:4d} {0[3]:5.2f} {0[4]:5.2f}\n'
.format(data))
fout.close()
if outall or kwargs.get('outcov'):
prody.writeArray(join(outdir, prefix + '_covariance' + ext),
anm.getCovariance(), delimiter=delim, format=format)
if outall or kwargs.get('outcc') or kwargs.get('outhm'):
cc = prody.calcCrossCorr(anm)
if outall or kwargs.get('outcc'):
prody.writeArray(join(outdir, prefix +
'_cross-correlations' + ext),
cc, delimiter=delim, format=format)
if outall or kwargs.get('outhm'):
prody.writeHeatmap(join(outdir, prefix + '_cross-correlations.hm'),
cc, resnum=select.getResnums(),
xlabel='Residue', ylabel='Residue',
title=anm.getTitle() + ' cross-correlations')
if outall or kwargs.get('hessian'):
prody.writeArray(join(outdir, prefix + '_hessian'+ext),
anm.getHessian(), delimiter=delim, format=format)
if outall or kwargs.get('kirchhoff'):
prody.writeArray(join(outdir, prefix + '_kirchhoff'+ext),
anm.getKirchhoff(), delimiter=delim, format=format)
if outall or kwargs.get('outsf'):
prody.writeArray(join(outdir, prefix + '_sqflucts'+ext),
prody.calcSqFlucts(anm), delimiter=delim,
format=format)
figall = kwargs.get('figall')
cc = kwargs.get('figcc')
sf = kwargs.get('figsf')
bf = kwargs.get('figbeta')
cm = kwargs.get('figcmap')
if figall or cc or sf or bf or cm:
try:
import matplotlib.pyplot as plt
except ImportError:
LOGGER.warning('Matplotlib could not be imported. '
'Figures are not saved.')
else:
prody.SETTINGS['auto_show'] = False
LOGGER.info('Saving graphical output.')
format = kwargs.get('figformat')
width = kwargs.get('figwidth')
height = kwargs.get('figheight')
dpi = kwargs.get('figdpi')
format = format.lower()
if figall or cc:
plt.figure(figsize=(width, height))
prody.showCrossCorr(anm)
plt.savefig(join(outdir, prefix + '_cc.'+format),
dpi=dpi, format=format)
plt.close('all')
if figall or cm:
plt.figure(figsize=(width, height))
prody.showContactMap(anm)
plt.savefig(join(outdir, prefix + '_cm.'+format),
dpi=dpi, format=format)
plt.close('all')
if figall or sf:
plt.figure(figsize=(width, height))
prody.showSqFlucts(anm)
plt.savefig(join(outdir, prefix + '_sf.'+format),
dpi=dpi, format=format)
plt.close('all')
if figall or bf:
plt.figure(figsize=(width, height))
bexp = select.getBetas()
bcal = prody.calcTempFactors(anm, select)
plt.plot(bexp, label='Experimental')
plt.plot(bcal, label=('Theoretical (R={0:.2f})'
.format(np.corrcoef(bcal, bexp)[0,1])))
plt.legend(prop={'size': 10})
plt.xlabel('Node index')
plt.ylabel('Experimental B-factors')
plt.title(pdb.getTitle() + ' B-factors')
plt.savefig(join(outdir, prefix + '_bf.'+format),
dpi=dpi, format=format)
plt.close('all')
def calcANMfeatures(self,
chain='all',
env='chain',
ANM_PRS=True,
stiffness=True,
MBS=False):
assert env in ['chain', 'reduced', 'sliced']
for k in ANM_PRS, stiffness, MBS:
assert type(k) is bool
# list of features to be computed
features = ['ANM_MSF-' + env]
if ANM_PRS:
features += ['ANM_effectiveness-' + env, 'ANM_sensitivity-' + env]
if MBS:
features += ['MBS-' + env]
if stiffness:
features += ['stiffness-' + env]
# compute features (if not precomputed)
if chain == 'all':
chain_list = self.chids
else:
chain_list = [
chain,
]
for chID in chain_list:
d = self.feats[chID]
if all([f in d for f in features]):
continue
try:
anm = self.calcANM(chID, env=env)
except Exception as e:
if (isinstance(e, MemoryError)):
msg = 'MemoryError'
else:
msg = str(e)
for f in features:
d[f] = msg
LOGGER.warn(msg)
continue
key_msf = 'ANM_MSF-' + env
if key_msf not in d:
try:
d[key_msf] = calcSqFlucts(anm)
except Exception as e:
msg = str(e)
d[key_msf] = msg
LOGGER.warn(msg)
key_eff = 'ANM_effectiveness-' + env
if key_eff in features and key_eff not in d:
key_sns = 'ANM_sensitivity-' + env
try:
prs_mtrx, eff, sns = calcPerturbResponse(anm)
d[key_eff] = eff
d[key_sns] = sns
except Exception as e:
msg = str(e)
d[key_eff] = msg
d[key_sns] = msg
LOGGER.warn(msg)
key_mbs = 'MBS-' + env
if key_mbs in features and key_mbs not in d:
try:
pdb = self.getPDB()
ca = pdb[chID].ca
d[key_mbs] = calcMBS(anm, ca, cutoff=15.)
except Exception as e:
msg = str(e)
d[key_mbs] = msg
LOGGER.warn(msg)
key_stf = 'stiffness-' + env
if key_stf in features and key_stf not in d:
try:
pdb = self.getPDB()
ca = pdb[chID].ca
stiff_mtrx = calcMechStiff(anm, ca)
d[key_stf] = np.mean(stiff_mtrx, axis=0)
except Exception as e:
msg = str(e)
d[key_stf] = msg
LOGGER.warn(msg)
return
def prody_pca(coords, **kwargs):
"""Perform PCA calculations for PDB or DCD format *coords* file.
"""
for key in DEFAULTS:
if not key in kwargs:
kwargs[key] = DEFAULTS[key]
from os.path import isdir, splitext, join
outdir = kwargs.get('outdir')
if not isdir(outdir):
raise IOError('{0} is not a valid path'.format(repr(outdir)))
import prody
LOGGER = prody.LOGGER
prefix = kwargs.get('prefix')
nmodes = kwargs.get('nmodes')
selstr = kwargs.get('select')
ext = splitext(coords)[1].lower()
if ext == '.gz':
ext = splitext(coords[:-3])[1].lower()
if ext == '.dcd':
pdb = kwargs.get('psf') or kwargs.get('pdb')
if pdb:
if splitext(pdb)[1].lower() == '.psf':
pdb = prody.parsePSF(pdb)
else:
pdb = prody.parsePDB(pdb)
dcd = prody.DCDFile(coords)
if prefix == '_pca' or prefix == '_eda':
prefix = dcd.getTitle() + prefix
if len(dcd) < 2:
raise ValueError('DCD file must have multiple frames')
if pdb:
if pdb.numAtoms() == dcd.numAtoms():
select = pdb.select(selstr)
dcd.setAtoms(select)
LOGGER.info('{0} atoms are selected for calculations.'
.format(len(select)))
else:
select = pdb.select(selstr)
if select.numAtoms() != dcd.numAtoms():
raise ValueError('number of selected atoms ({0}) does '
'not match number of atoms in the DCD '
'file ({1})'.format(select.numAtoms(),
dcd.numAtoms()))
if pdb.numCoordsets():
dcd.setCoords(select.getCoords())
else:
select = prody.AtomGroup()
select.setCoords(dcd.getCoords())
pca = prody.PCA(dcd.getTitle())
if len(dcd) > 1000:
pca.buildCovariance(dcd, aligned=kwargs.get('aligned'))
pca.calcModes(nmodes)
ensemble = dcd
else:
ensemble = dcd[:]
if not kwargs.get('aligned'):
ensemble.iterpose()
pca.performSVD(ensemble)
else:
pdb = prody.parsePDB(coords)
if pdb.numCoordsets() < 2:
raise ValueError('PDB file must contain multiple models')
if prefix == '_pca' or prefix == '_eda':
prefix = pdb.getTitle() + prefix
select = pdb.select(selstr)
LOGGER.info('{0} atoms are selected for calculations.'
.format(len(select)))
if select is None:
raise ValueError('selection {0} do not match any atoms'
.format(repr(selstr)))
LOGGER.info('{0} atoms will be used for PCA calculations.'
.format(len(select)))
ensemble = prody.Ensemble(select)
pca = prody.PCA(pdb.getTitle())
if not kwargs.get('aligned'):
ensemble.iterpose()
pca.performSVD(ensemble)
LOGGER.info('Writing numerical output.')
if kwargs.get('outnpz'):
prody.saveModel(pca, join(outdir, prefix))
prody.writeNMD(join(outdir, prefix + '.nmd'), pca[:nmodes], select)
extend = kwargs.get('extend')
if extend:
if pdb:
if extend == 'all':
extended = prody.extendModel(pca[:nmodes], select, pdb)
else:
extended = prody.extendModel(pca[:nmodes], select,
select | pdb.bb)
prody.writeNMD(join(outdir, prefix + '_extended_' +
extend + '.nmd'), *extended)
else:
prody.LOGGER.warn('Model could not be extended, provide a PDB or '
'PSF file.')
outall = kwargs.get('outall')
delim = kwargs.get('numdelim')
ext = kwargs.get('numext')
format = kwargs.get('numformat')
if outall or kwargs.get('outeig'):
prody.writeArray(join(outdir, prefix + '_evectors'+ext),
pca.getArray(), delimiter=delim, format=format)
prody.writeArray(join(outdir, prefix + '_evalues'+ext),
pca.getEigvals(), delimiter=delim, format=format)
if outall or kwargs.get('outcov'):
prody.writeArray(join(outdir, prefix + '_covariance'+ext),
pca.getCovariance(), delimiter=delim, format=format)
if outall or kwargs.get('outcc') or kwargs.get('outhm'):
cc = prody.calcCrossCorr(pca)
if outall or kwargs.get('outcc'):
prody.writeArray(join(outdir, prefix + '_cross-correlations' +
ext), cc, delimiter=delim, format=format)
if outall or kwargs.get('outhm'):
resnums = select.getResnums()
hmargs = {} if resnums is None else {'resnums': resnums}
prody.writeHeatmap(join(outdir, prefix + '_cross-correlations.hm'),
cc, xlabel='Residue', ylabel='Residue',
title=pca.getTitle() + ' cross-correlations',
**hmargs)
if outall or kwargs.get('outsf'):
prody.writeArray(join(outdir, prefix + '_sqfluct'+ext),
prody.calcSqFlucts(pca), delimiter=delim,
format=format)
if outall or kwargs.get('outproj'):
prody.writeArray(join(outdir, prefix + '_proj'+ext),
prody.calcProjection(ensemble, pca), delimiter=delim,
format=format)
figall = kwargs.get('figall')
cc = kwargs.get('figcc')
sf = kwargs.get('figsf')
sp = kwargs.get('figproj')
if figall or cc or sf or sp:
try:
import matplotlib.pyplot as plt
except ImportError:
LOGGER.warning('Matplotlib could not be imported. '
'Figures are not saved.')
else:
prody.SETTINGS['auto_show'] = False
LOGGER.info('Saving graphical output.')
format = kwargs.get('figformat')
width = kwargs.get('figwidth')
height = kwargs.get('figheight')
dpi = kwargs.get('figdpi')
format = format.lower()
if figall or cc:
plt.figure(figsize=(width, height))
prody.showCrossCorr(pca)
plt.savefig(join(outdir, prefix + '_cc.'+format),
dpi=dpi, format=format)
plt.close('all')
if figall or sf:
plt.figure(figsize=(width, height))
prody.showSqFlucts(pca)
plt.savefig(join(outdir, prefix + '_sf.'+format),
dpi=dpi, format=format)
plt.close('all')
if figall or sp:
indices = []
for item in sp.split():
try:
if '-' in item:
item = item.split('-')
if len(item) == 2:
indices.append(list(range(int(item[0])-1,
int(item[1]))))
elif ',' in item:
indices.append([int(i)-1 for i in item.split(',')])
else:
indices.append(int(item)-1)
except:
pass
for index in indices:
plt.figure(figsize=(width, height))
prody.showProjection(ensemble, pca[index])
if isinstance(index, int):
index = [index]
index = [str(i+1) for i in index]
plt.savefig(join(outdir, prefix + '_proj_' +
'_'.join(index) + '.' + format),
dpi=dpi, format=format)
plt.close('all')
def iENM_stochastic( pdb_obj, prefix, selstr='calpha', dr=1, nsteps=1, const=1, cutoff=10, selmode=-1, path_length=0.5 ): import prody as pd import pylab as pb import numpy as np import random as random pdb = pdb_obj.select( selstr ) natoms = pdb.numAtoms() dcd = np.zeros( ( natoms*3, 1 ) ) dcd = pdb.getCoords().reshape( natoms*3 ) ensemble = pd.Ensemble() ensemble.setCoords(dcd.reshape(natoms,3)) filename_pdb = '%s.pdb' % prefix filename_dcd = '%s.dcd' % prefix pd.writePDB( filename_pdb, pdb ) for i in xrange( 1, nsteps+1 ): print 'Calculating coordinates at step %d\n' % i # Load coordinates from previous step pdb.setCoords( dcd.reshape(natoms,3) ) # GNM analysis gnm = pd.GNM() gnm.buildKirchhoff( pdb, cutoff=cutoff, gamma=const ) K = gnm.getKirchhoff() N = np.shape(K)[1] gnm.calcModes( n_modes=N,zeros=False,turbo=True ) if selmode == -1: fluct = pd.calcSqFlucts( gnm ) / max( pd.calcSqFlucts( gnm ) ) if selmode != -1: fluct = pd.calcSqFlucts( gnm[selmode] ) / max( pd.calcSqFlucts( gnm[selmode] ) ) # ANM analysis anm = pd.ANM() anm.buildHessian( pdb, cutoff=cutoff, gamma=const ) H = anm.getHessian() N = (np.shape(H)[1] / 3) anm.calcModes( n_modes=N,zeros=False,turbo=True ) eVec = anm.getEigvecs() eVal = anm.getEigvals() # Normalize vectors for i in xrange (0, np.size(eVec[1]) ): eVec[:,i] = norm_vec( eVec[:,i], natoms ) # Randomly sample gaussian distribution of vectors gVec = gauss_vec(eVec) # Weight motion along gVec by gnm square fluctuations rvec = gVec rmag = np.zeros((natoms,3)) rmag[:,0] = fluct rmag[:,1] = rmag[:,0] rmag[:,2] = rmag[:,0] rmag = rmag.reshape(natoms*3) rmag = ((-1)**np.random.random_integers(0,1,np.shape(rmag))) * path_length * rmag # rmag = ((-1)**random.randint(0,1)) * rmag rmag = rmag.reshape(natoms*3) rvec = rvec.reshape(natoms*3) rvec = np.multiply(rvec,rmag) # Predict new coords dcd = rvec.reshape(natoms*3) + pdb.getCoords().reshape(natoms*3) ensemble.addCoordset(dcd.reshape(natoms,3)) pd.writeDCD( filename_dcd, ensemble ) return ( pdb, ensemble )
def prody_pca(opt):
"""Perform PCA calculations based on command line arguments."""
outdir = opt.outdir
if not os.path.isdir(outdir):
opt.subparser.error('{0:s} is not a valid path'.format(outdir))
import prody
LOGGER = prody.LOGGER
coords = opt.coords
prefix = opt.prefix
nmodes, selstr = opt.nmodes, opt.select
if os.path.splitext(coords)[1].lower() == '.dcd':
ag = opt.psf or opt.pdb
if ag:
if os.path.splitext(ag)[1].lower() == '.psf':
ag = prody.parsePSF(ag)
else:
ag = prody.parsePDB(ag)
dcd = prody.DCDFile(opt.coords)
if len(dcd) < 2:
opt.subparser("DCD file must contain multiple frames.")
if ag:
dcd.setAtomGroup(ag)
select = dcd.select(selstr)
LOGGER.info('{0:d} atoms are selected for calculations.'
.format(len(select)))
else:
select = prody.AtomGroup()
select.setCoords(dcd.getCoords())
pca = prody.PCA(dcd.getTitle())
if len(dcd) > 1000:
pca.buildCovariance(dcd)
pca.calcModes(dcd)
else:
pca.performSVD(dcd[:])
else:
pdb = prody.parsePDB(opt.coords)
if pdb.numCoordsets() < 2:
opt.subparser("PDB file must contain multiple models.")
if prefix == '_pca':
prefix = pdb.getTitle() + '_pca'
select = pdb.select(selstr)
LOGGER.info('{0:d} atoms are selected for calculations.'
.format(len(select)))
if select is None:
opt.subparser('Selection "{0:s}" do not match any atoms.'
.format(selstr))
LOGGER.info('{0:d} atoms will be used for PCA calculations.'
.format(len(select)))
ensemble = prody.Ensemble(select)
pca = prody.PCA(pdb.getTitle())
ensemble.iterpose()
pca.performSVD(ensemble)
LOGGER.info('Writing numerical output.')
if opt.npz:
prody.saveModel(pca)
prody.writeNMD(os.path.join(outdir, prefix + '.nmd'), pca[:nmodes], select)
outall = opt.all
delim, ext, format = opt.delim, opt.ext, opt.numformat
if outall or opt.eigen:
prody.writeArray(os.path.join(outdir, prefix + '_evectors'+ext),
pca.getArray(), delimiter=delim, format=format)
prody.writeArray(os.path.join(outdir, prefix + '_evalues'+ext),
pca.getEigenvalues(), delimiter=delim, format=format)
if outall or opt.covar:
prody.writeArray(os.path.join(outdir, prefix + '_covariance'+ext),
pca.getCovariance(), delimiter=delim, format=format)
if outall or opt.ccorr:
prody.writeArray(os.path.join(outdir, prefix + '_cross-correlations' +
ext), prody.calcCrossCorr(pca),
delimiter=delim, format=format)
if outall or opt.sqflucts:
prody.writeArray(os.path.join(outdir, prefix + '_sqfluct'+ext),
prody.calcSqFlucts(pca), delimiter=delim,
format=format)
if outall or opt.proj:
prody.writeArray(os.path.join(outdir, prefix + '_proj'+ext),
prody.calcProjection(ensemble, pca), delimiter=delim,
format=format)
figall, cc, sf, sp = opt.figures, opt.cc, opt.sf, opt.sp
if figall or cc or sf or sp:
try:
import matplotlib.pyplot as plt
except ImportError:
LOGGER.warning('Matplotlib could not be imported. '
'Figures are not saved.')
else:
LOGGER.info('Saving graphical output.')
format, width, height, dpi = \
opt.figformat, opt.width, opt.height, opt.dpi
format = format.lower()
if figall or cc:
plt.figure(figsize=(width, height))
prody.showCrossCorr(pca)
plt.savefig(os.path.join(outdir, prefix + '_cc.'+format),
dpi=dpi, format=format)
plt.close('all')
if figall or sf:
plt.figure(figsize=(width, height))
prody.showSqFlucts(pca)
plt.savefig(os.path.join(outdir, prefix + '_sf.'+format),
dpi=dpi, format=format)
plt.close('all')
if figall or sp:
indices = []
for item in sp.split():
try:
if '-' in item:
item = item.split('-')
if len(item) == 2:
indices.append(range(int(item[0])-1,
int(item[1])))
elif ',' in item:
indices.append([int(i)-1 for i in item.split(',')])
else:
indices.append(int(item)-1)
except:
pass
for index in indices:
plt.figure(figsize=(width, height))
prody.showProjection(ensemble, pca[index])
if isinstance(index, int):
index = [index]
index = [str(i+1) for i in index]
plt.savefig(os.path.join(outdir, prefix + '_proj_' +
'_'.join(index) + '.' + format),
dpi=dpi, format=format)
plt.close('all')
def iENM_deformationE( pdb_obj, ref_pdb_obj, prefix, selstr='calpha', dr=1, nsteps=1, const=1, cutoff=10, path_length=1 ): import prody as pd import pylab as pb import numpy as np import random as random pdb = pdb_obj.select( selstr ) natoms = pdb.numAtoms() dcd = np.zeros( ( natoms*3, 1 ) ) dcd = pdb.getCoords().reshape( natoms*3 ) ensemble = pd.Ensemble() ensemble.setCoords(dcd.reshape(natoms,3)) filename_pdb = '%s.pdb' % prefix filename_dcd = '%s.dcd' % prefix pd.writePDB( filename_pdb, pdb ) ref = ref_pdb_obj.select( selstr ) for i in xrange( 1, nsteps+1 ): print 'Calculating coordinates at step %d\n' % i # Load coordinates from previous step pdb.setCoords( dcd.reshape((natoms,3)) ) # ANM analysis anm = pd.ANM() anm.buildHessian( pdb, cutoff=cutoff, gamma=const ) H = anm.getHessian() # N = (np.shape(H)[1] / 3) if np.size(selmode) > 1: N = max(selmode) + 1 if np.size(selmode) == 1: N = selmode + 1 anm.calcModes( n_modes=N,zeros=False,turbo=True ) eVec = anm.getEigvecs() eVal = anm.getEigvals() # Normalize eigenvectors and generate gaussian if needed if np.size(selmode) > 1: rvec = np.zeros((np.shape(eVec)[0],np.size(selmode))) for i in xrange ( 0, np.size(eVec[1]) ): rvec[:,i] = norm_vec( eVec[:,selmode[i]], natoms ) rvec = gauss_vec(rvec) if np.size(selmode) == 1: if selmode == 0: rvec = norm_vec( eVec, natoms ) if selmode > 0: rvec = norm_vec( eVec[:,selmode], natoms ) # if np.size(selmode) > 1: # for i in xrange (0, np.size(eVec[1]) ): # rvec = norm_vec( eVec[:,selmode[i]], natoms ) # rvec = gauss_vec(rvec[:,selmode]) # if np.size(selmode) == 1: # rvec = norm_vec( eVec[:,selmode], natoms ) rmag = np.zeros((natoms,3)) fluct = pd.calcSqFlucts(anm[selmode]) fluct = fluct / max(fluct) rmag[:,0] = fluct rmag[:,1] = fluct rmag[:,2] = fluct rmag = rmag.reshape(natoms*3) # rmag = ((-1)**random.randint(0,1)) * path_length * rmag # Commented out to look at positive eigenvector! rmag = path_length * rmag rmag = rmag.reshape(natoms*3) rvec = rvec.reshape(natoms*3) rvec = np.multiply(rvec,rmag) # Predict new coords dcd = rvec + pdb.getCoords().reshape( natoms*3 ) ensemble.addCoordset(dcd.reshape(natoms,3)) pd.writeDCD( filename_dcd, ensemble )
