def unselResidues(): from chimera.selection import currentResidues selResidues = currentResidues(asDict=True) unsel = [] for m in chimera.openModels.list(modelTypes=[chimera.Molecule]): unsel.extend(filter(lambda r: r not in selResidues, m.residues)) return unsel
def unselResidues():
from chimera.selection import currentResidues
selResidues = currentResidues(asDict=True)
unsel = []
for m in chimera.openModels.list(modelTypes=[chimera.Molecule]):
unsel.extend(filter(lambda r: r not in selResidues, m.residues))
return unsel
def standardize_terminal_protein_residues(receptor_pdb, molID="#0"):
"""
To prevent errors like 'ValueError: Cannot determine GAFF type for :11.A@HD14 (etc.)' raised by
initiateAddions(), originating from termini capped by Shrodinger's Maestro or incomplete or missing protein
residues, mutate the terminal residues to their original aa type (applies to every protein chain).
"""
print("Standardizing protein's terminal residues.")
rc("sel %s & protein" % molID)
chaindID_resids = defaultdict(list)
for r in currentResidues():
chaindID_resids[r.id.chainId].append((r.id.position, r.type))
for chainID in chaindID_resids.keys():
chaindID_resids[chainID].sort(key=itemgetter(0))
# NOTE: the side-chain mutations (swapaa) were not necessary in the proteins tested so far.
# rc("swapaa %s %s:%i.%s" %
# (chaindID_resids[chainID][0][1], molID, chaindID_resids[chainID][0][0], str(chainID))) # N-term
rc("del element.H & %s:%i.%s" %
(molID, chaindID_resids[chainID][0][0], str(chainID)))
# rc("swapaa %s %s:%i.%s" %
# (chaindID_resids[chainID][-1][1], molID, chaindID_resids[chainID][-1][0], str(chainID))) # C-term
rc("del element.H & %s:%i.%s" %
(molID, chaindID_resids[chainID][-1][0], str(chainID)))
# Only if you write and load the PDB then Chimera will reset the valence of the N-terminal amide and
# thus will not again the H1,H2,H3 which cause the error.
rc("write format pdb #0 " + receptor_pdb.replace(".pdb", "_tmp.pdb"))
rc("del #0")
rc("open %s %s" % (molID, receptor_pdb.replace(".pdb", "_tmp.pdb")))
os.remove(receptor_pdb.replace(".pdb", "_tmp.pdb"))
def NDBColors(self):
from chimera import selection
if selection.currentEmpty():
import Midas
residues = Midas._selectedResidues('#')
else:
residues = selection.currentResidues()
NA.NDBColors(residues)
def apply(self, restrict=False):
if restrict:
from chimera import selection
rList = selection.currentResidues()
else:
rList = []
if rList:
for r in rList:
self.setResidue(r)
else:
for m in chimera.openModels.list(modelTypes=[chimera.Molecule]):
for r in m.residues:
self.setResidue(r)
def apply(self, restrict=False): if restrict: from chimera import selection rList = selection.currentResidues() else: rList = [] if rList: for r in rList: self.setResidue(r) else: for m in chimera.openModels.list( modelTypes=[chimera.Molecule]): for r in m.residues: self.setResidue(r)
def Apply(self, restrict):
if self.currentRRC:
rrc = self.currentRRC
else:
rrc = RRC(None, self.guide, self.plane, self.planeNormal,
self.isNucleic, self.mainchain)
if restrict:
from chimera import selection
rList = selection.currentResidues()
else:
rList = []
if rList:
for r in rList:
rrc.setResidue(r)
else:
for m in chimera.openModels.list(modelTypes=[chimera.Molecule]):
for r in m.residues:
rrc.setResidue(r)
def _getMolecules(self, assignsel, usesel, restrict): if assignsel or usesel: sel = selection.currentMolecules() if usesel: for r in selection.currentResidues(): restrict[r] = True for a in selection.currentAtoms(): restrict[a] = True for m in sel: restrict[m] = True if assignsel: molecules = sel else: molecules = [] if not molecules: molecules = chimera.openModels.list( modelTypes=[chimera.Molecule]) return molecules
def Apply(self, restrict):
if self.currentXS:
xs = self.currentXS
else:
xs = XS(None, self.points, self.smoothacross, self.smoothalong,
self.closed, self.divisions)
if restrict:
from chimera import selection
rList = selection.currentResidues()
else:
rList = []
if rList:
for r in rList:
xs.setResidue(r)
else:
for m in chimera.openModels.list(modelTypes=[chimera.Molecule]):
for r in m.residues:
xs.setResidue(r)
def showChimeraSelection(self):
selRegion = self.getRegion(SEL_REGION_NAME, create=1,
fill=self.seqCanvas.mav.prefs[SEL_REGION_INTERIOR],
outline=self.seqCanvas.mav.prefs[SEL_REGION_BORDER])
selRegion.clear()
resDict = {}
for res in currentResidues():
resDict[res] = 1
blocks = []
for aseq in self.seqCanvas.seqs:
try:
matchMaps = aseq.matchMaps
except AttributeError:
continue
for matchMap in matchMaps.values():
start = None
end = None
for i in range(len(aseq.ungapped())):
if i in matchMap \
and matchMap[i] in resDict:
if start is not None:
end = i
else:
end = start = i
else:
if start is not None:
blocks.append([aseq,
aseq, aseq. \
ungapped2gapped(start
), aseq. \
ungapped2gapped(end)])
start = end = None
if start is not None:
blocks.append([aseq, aseq,
aseq.ungapped2gapped(start),
aseq.ungapped2gapped(end)])
if blocks and self._firstChimeraShow:
self._firstChimeraShow = False
self.seqCanvas.mav.status(
"Chimera selection region displayed.\n"
"Preferences..Regions controls this display.\n")
selRegion.addBlocks(blocks)
self.raiseRegion(selRegion)
def Apply(self, restrict): if self.currentXS: xs = self.currentXS else: xs = XS(None, self.points, self.smoothacross, self.smoothalong, self.closed, self.divisions) if restrict: from chimera import selection rList = selection.currentResidues() else: rList = [] if rList: for r in rList: xs.setResidue(r) else: for m in chimera.openModels.list( modelTypes=[chimera.Molecule]): for r in m.residues: xs.setResidue(r)
def Apply(self, restrict): if self.currentRRC: rrc = self.currentRRC else: rrc = RRC(None, self.guide, self.plane, self.planeNormal, self.isNucleic, self.mainchain) if restrict: from chimera import selection rList = selection.currentResidues() else: rList = [] if rList: for r in rList: rrc.setResidue(r) else: for m in chimera.openModels.list( modelTypes=[chimera.Molecule]): for r in m.residues: rrc.setResidue(r)
def selResidues(noneReturnsAll=True, implied=False, create=False):
residues = selection.currentResidues()
extendSelection(residues, 'residues', noneReturnsAll, implied, create)
return residues
def nw(s1, s2, scoreMatch=10, scoreMismatch=-3, scoreGap=0, scoreGapOpen=-40,
gapChar=".", returnSeqs=False, scoreMatrix=None,
similarityMatrix=None, frequencyMatrix=None,
endsAreGaps=False, ssMatrix=None, ssFraction=0.9,
gapOpenHelix=None, gapOpenStrand=None,
gapOpenOther=None, debug=False):
# if 'scoreMatrix', 'similarityMatrix', or 'frequencyMatrix' is
# provided, then 'scoreMatch' and 'scoreMismatch' are ignored and
# the matrix is used to evaluate matching between the sequences.
# 'scoreMatrix' should be a two-dimensional array of size
# len(s1) x len(s2). 'similarityMatrix' should be a dictionary
# keyed with two-tuples of residue types. 'frequencyMatrix' should
# be a list of length s2 of dictionaries, keyed by residue type.
#
# if 'ssFraction' is not None/False, then 'ssMatrix' should be a 3x3
# matrix keyed with 2-tuples of secondary structure types ('H': helix,
# 'S': strand, 'O': other). The score will be a mixture of the
# ss/similarity matrix scores weighted by the ssFraction
# [ssFraction * ss score + (1 - ssFraction) * similarity score]
#
# if 'gapOpenHelix/Strand/Other' is not None and 'ssFraction' is not
# None/False, then scoreGapOpen is ignored when an intra-helix/
# intra-strand/other gap is opened and the appropriate penalty
# is applied instead
#
# if 'returnSeqs' is True, then instead of returning a match list
# (a list of two-tuples) as the second value, a two-tuple of gapped
# Sequences will be returned. In both cases, the first return value
# is the match score.
m = []
bt = []
for i1 in range(len(s1) + 1):
m.append((len(s2) + 1) * [ 0 ])
bt.append((len(s2) + 1) * [None])
bt[i1][0] = 1
if endsAreGaps and i1 > 0:
m[i1][0] = scoreGapOpen + i1 * scoreGap
for i2 in range(len(s2) + 1):
bt[0][i2] = 2
if endsAreGaps and i2 > 0:
m[0][i2] = scoreGapOpen * i2 * scoreGap
if similarityMatrix is not None:
evaluate = lambda i1, i2: similarityMatrix[(s1[i1], s2[i2])]
elif scoreMatrix is not None:
evaluate = lambda i1, i2: scoreMatrix[i1][i2]
elif frequencyMatrix is not None:
evaluate = lambda i1, i2: frequencyMatrix[i2][s1[i1]]
else:
def evaluate(i1, i2):
if s1[i1] == s2[i2]:
return scoreMatch
return scoreMismatch
doingSS = ssFraction is not None and ssFraction is not False \
and ssMatrix is not None
if doingSS:
prevEval = evaluate
simFraction = 1.0 - ssFraction
def ssEval(i1, i2):
if hasattr(s1, 'ssFreqs'):
freqs1 = s1.ssFreqs[i1]
else:
freqs1 = {s1.ssType(i1): 1.0}
if hasattr(s2, 'ssFreqs'):
freqs2 = s2.ssFreqs[i2]
else:
freqs2 = {s2.ssType(i2): 1.0}
val = 0.0
for ss1, freq1 in freqs1.items():
if ss1 == None:
continue
for ss2, freq2 in freqs2.items():
if ss2 == None:
continue
val += freq1 * freq2 * ssMatrix[(ss1,
ss2)]
return val
evaluate = lambda i1, i2: ssFraction * ssEval(i1, i2) + \
simFraction * prevEval(i1, i2)
# precompute appropriate gap-open penalties
gapOpen1 = [scoreGapOpen] * (len(s1)+1)
gapOpen2 = [scoreGapOpen] * (len(s2)+1)
if endsAreGaps:
if gapOpenOther is not None:
gapOpen1[0] = gapOpen2[0] = gapOpenOther
else:
gapOpen1[0] = gapOpen2[0] = 0
if doingSS and gapOpenOther != None:
for seq, gapOpens in [(s1, gapOpen1), (s2, gapOpen2)]:
if hasattr(seq, 'gapFreqs'):
for i, gapFreq in enumerate(seq.gapFreqs):
gapOpens[i+1] = \
gapFreq['H'] * gapOpenHelix + \
gapFreq['S'] * gapOpenStrand + \
gapFreq['O'] * gapOpenOther
else:
ssTypes = [seq.ssType(i)
for i in range(len(seq))]
for i, ss in enumerate(ssTypes[:-1]):
nextSS = ssTypes[i+1]
if ss == nextSS and ss == 'H':
gapOpens[i+1] = gapOpenHelix
elif ss == nextSS and ss == 'S':
gapOpens[i+1] = gapOpenStrand
else:
gapOpens[i+1] = gapOpenOther
colGapStarts = [0] * len(s2) # don't care about column zero
for i1 in range(len(s1)):
rowGapPos = 0
for i2 in range(len(s2)):
best = m[i1][i2] + evaluate(i1, i2)
btType = 0
if i2 + 1 < len(s2) or endsAreGaps:
colGapPos = colGapStarts[i2]
skipSize = i1 + 1 - colGapPos
if hasattr(s1, "occupancy"):
totOcc = 0.0
for i in range(colGapPos, i1+1):
totOcc += s1.occupancy[i]
colSkipVal = totOcc * scoreGap
else:
colSkipVal = skipSize * scoreGap
baseColGapVal = m[colGapPos][i2+1] + colSkipVal
skip = baseColGapVal + gapOpen2[i2+1]
else:
skipSize = 1
colSkipVal = 0
skip = m[i1][i2+1]
if skip > best:
best = skip
btType = skipSize
if i1 + 1 < len(s1) or endsAreGaps:
skipSize = i2 + 1 - rowGapPos
if hasattr(s2, "occupancy"):
totOcc = 0.0
for i in range(rowGapPos, i2+1):
totOcc += s2.occupancy[i]
rowSkipVal = totOcc * scoreGap
else:
rowSkipVal = skipSize * scoreGap
baseRowGapVal = m[i1+1][rowGapPos] + rowSkipVal
skip = baseRowGapVal + gapOpen1[i1+1]
else:
skipSize = 1
rowSkipVal = 0
skip = m[i1+1][i2]
if skip > best:
best = skip
btType = 0 - skipSize
m[i1+1][i2+1] = best
bt[i1+1][i2+1] = btType
if btType >= 0:
# not gapping the row
if best > baseRowGapVal:
rowGapPos = i2 + 1
if btType <= 0:
# not gapping the column
if best > baseColGapVal:
colGapStarts[i2] = i1 + 1
if debug:
from chimera.selection import currentResidues
cr = currentResidues(asDict=True)
if cr:
for fileName, matrix in [("scores", m), ("trace", bt)]:
out = open("/home/socr/a/pett/rm/" + fileName,
"w")
print>>out, " ",
for i2, r2 in enumerate(s2.residues):
if r2 not in cr:
continue
print>>out, "%5d" % i2,
print>>out
print>>out, " ",
for i2, r2 in enumerate(s2.residues):
if r2 not in cr:
continue
print>>out, "%5s" % s2[i2],
print>>out
for i1, r1 in enumerate(s1.residues):
if r1 not in cr:
continue
print>>out, "%3d" % i1, s1[i1],
for i2, r2 in enumerate(s2.residues):
if r2 not in cr:
continue
print>>out, "%5g" % (
matrix[i1+1][i2+1]),
print>>out
out.close()
i1 = len(s1)
i2 = len(s2)
matchList = []
while i1 > 0 and i2 > 0:
btType = bt[i1][i2]
if btType == 0:
matchList.append((i1-1, i2-1))
i1 = i1 - 1
i2 = i2 - 1
elif btType > 0:
i1 = i1 - btType
else:
i2 = i2 + btType
if returnSeqs:
return m[len(s1)][len(s2)], matches2gappedSeqs(matchList,
s1, s2, gapChar=gapChar)
return m[len(s1)][len(s2)], matchList
def makeAllMeshes(proID, jobID, peptideLen=40, pocketDist=5.0):
global curResID
global curProID
global meshDir
global pocketMap
global resCount
global curJobID
#Insure input is a 4 character protein ID
proID = proID[0:4]
#Create a new directory for the protein
meshDir = proID + 'Mesh'
curProID = proID
curJobID = jobID
# os.mkdir(meshDir)
# os.system("chmod 777 " + meshDir)
try:
#Fetch protein from the PDB
runCommand('open ' + proID)
#Initial prep for pocket search
runCommand('delete solvent')
runCommand('delete :HOH')
#Open protein model
model = openModels.list(modelTypes=[Molecule])[0]
RES_LIST = model.residues
resCount = len(RES_LIST)
#Pocket finding procedure
chains = []
counts = []
index = 0
print 'Counting Chain Lengths................'
for res in RES_LIST:
curChain = str(res.id.chainId)
if len(chains) == 0:
chains.append(curChain)
counts.append(1)
if curChain != chains[index]:
index = index + 1
chains.append(curChain)
counts.append(1)
else:
counts[index] = counts[index] + 1
print 'Finding Peptide Ligand Pockets................'
for i, chain in enumerate(chains):
if counts[i] < peptideLen: #Less than X amino acids = peptide
runCommand('~select')
runCommand('select ligand & :.' +
chain) #Check if actually ligand
runCommand('select selected zr<' +
str(pocketDist)) #Select pocket
runCommand('~select :/isHet zr<' +
str(pocketDist)) #De-select ligand
runCommand('~select ligand')
residues = selection.currentResidues() #Store pocket in memory
for res in residues:
if res in pocketMap:
pocketMap[res].append('|' + chain)
else:
pocketMap[res] = [chain]
print 'Finding non-Peptide Ligand Pockets................'
for res in RES_LIST:
if res.isHet: #Check if "heterogenous residue" which includes ligands
curResID = str(res.id)
runCommand('~select')
runCommand('select :' + curResID +
' & ligand') #Check if actually ligand
runCommand('select selected zr<' +
str(pocketDist)) #Select pocket
runCommand('~select :' + curResID)
residues = selection.currentResidues() #Store pocket in memory
for pocketRes in residues:
if pocketRes in pocketMap:
pocketMap[pocketRes].append('|' + res.type + '-' +
curResID)
else:
pocketMap[pocketRes] = [res.type + '-' + curResID]
print 'Dock Prep................'
prep(openModels.list(modelTypes=[Molecule]))
runCommand('delete ligand')
print '\nRunning Delphi................'
runCommand('~select')
runCommand('write format pdb 0 ' + proID + curJobID +
'.pdb') #Write delphi-ready protein with hydrogens
#Create delphi parameter file
f = open('./param_' + curJobID, 'w')
f.write('in(pdb,file="' + proID + curJobID +
'.pdb")\n') #Input/Output files
f.write('in(siz,file="amber.siz")\n')
f.write('in(crg,file="amber.crg")\n')
f.write('out(phi,file="' + proID + curJobID + '.phi")\n')
f.write('indi=2\n') #Interior dielectric constant
f.write('exdi=80.0\n') #Exterior dielectric constant (water)
f.write('prbrad=1.4\n')
f.write('salt=0.15\n') #Salt concentration and radius
f.write('ionrad=2.0\n')
# f.write('nonit=20\n') #Non-linear iterations
f.close()
os.system("./DELPHI_2004_LINUX_v2/delphi_static param_" +
curJobID) #Run delphi
# print '\nExporting Mesh................'
runCommand('select #0')
runCommand('surface')
#runCommand('export format X3D ./' + proID + '.x3d') # Prints raw X3D file.
print '\nOpening Delphi Data................'
runCommand('open ' + proID + curJobID + '.phi')
runCommand('scolor #0 volume #1 cmap -1,red:0,white:1,blue;')
print '\nRemoving Temp Files................'
os.remove("./" + proID + curJobID + '.pdb')
os.remove("./" + proID + curJobID + '.phi')
os.remove('./param_' + curJobID)
print '\nDONE: ' + proID
except:
f = open("./" + 'ErrorLog' + curJobID + '.txt', 'w')
f.write(proID + ' error: ' + str(sys.exc_info()[0]) +
str(sys.exc_info()[1]) + '\n')
f.close()
raise
return
def makeAllMeshes(proID, jobID, peptideLen=40, pocketDist=5.0):
global curResID
global curProID
global meshDir
global pocketMap
global resCount
global curJobID
#Insure input is a 4 character protein ID
proID = proID[0:4]
#Create a new directory for the protein
meshDir = proID + 'Mesh'
curProID = proID
curJobID = jobID
# os.mkdir(meshDir)
# os.system("chmod 777 " + meshDir)
try:
#Fetch protein from the PDB
runCommand('open '+ proID)
#Initial prep for pocket search
runCommand('delete solvent')
runCommand('delete :HOH')
#Open protein model
model = openModels.list(modelTypes=[Molecule])[0]
RES_LIST = model.residues
resCount = len(RES_LIST)
#Pocket finding procedure
chains = []
counts = []
index = 0
print 'Counting Chain Lengths................'
for res in RES_LIST:
curChain = str(res.id.chainId)
if len(chains) == 0:
chains.append(curChain)
counts.append(1)
if curChain != chains[index]:
index = index + 1
chains.append(curChain)
counts.append(1)
else:
counts[index] = counts[index] + 1
print 'Finding Peptide Ligand Pockets................'
for i, chain in enumerate(chains):
if counts[i] < peptideLen: #Less than X amino acids = peptide
runCommand('~select')
runCommand('select ligand & :.' + chain) #Check if actually ligand
runCommand('select selected zr<' + str(pocketDist)) #Select pocket
runCommand('~select :/isHet zr<' + str(pocketDist)) #De-select ligand
runCommand('~select ligand')
residues = selection.currentResidues() #Store pocket in memory
for res in residues:
if res in pocketMap:
pocketMap[res].append('|' + chain)
else:
pocketMap[res] = [chain]
print 'Finding non-Peptide Ligand Pockets................'
for res in RES_LIST:
if res.isHet: #Check if "heterogenous residue" which includes ligands
curResID = str(res.id)
runCommand('~select')
runCommand('select :' + curResID + ' & ligand') #Check if actually ligand
runCommand('select selected zr<' + str(pocketDist)) #Select pocket
runCommand('~select :' + curResID)
residues = selection.currentResidues() #Store pocket in memory
for pocketRes in residues:
if pocketRes in pocketMap:
pocketMap[pocketRes].append('|' + res.type + '-' + curResID)
else:
pocketMap[pocketRes] = [res.type + '-' + curResID]
print 'Dock Prep................'
prep(openModels.list(modelTypes=[Molecule]))
runCommand('delete ligand')
print '\nRunning Delphi................'
runCommand('~select')
runCommand('write format pdb 0 '+ proID + curJobID + '.pdb') #Write delphi-ready protein with hydrogens
#Create delphi parameter file
f = open('./param_' + curJobID, 'w')
f.write('in(pdb,file="'+ proID + curJobID + '.pdb")\n') #Input/Output files
f.write('in(siz,file="amber.siz")\n')
f.write('in(crg,file="amber.crg")\n')
f.write('out(phi,file="'+ proID + curJobID + '.phi")\n')
f.write('indi=2\n') #Interior dielectric constant
f.write('exdi=80.0\n') #Exterior dielectric constant (water)
f.write('prbrad=1.4\n')
f.write('salt=0.15\n') #Salt concentration and radius
f.write('ionrad=2.0\n')
# f.write('nonit=20\n') #Non-linear iterations
f.close()
os.system("./DELPHI_2004_LINUX_v2/delphi_static param_" + curJobID) #Run delphi
# print '\nExporting Mesh................'
runCommand('select #0')
runCommand('surface')
#runCommand('export format X3D ./' + proID + '.x3d') # Prints raw X3D file.
print '\nOpening Delphi Data................'
runCommand('open '+ proID + curJobID + '.phi')
runCommand('scolor #0 volume #1 cmap -1,red:0,white:1,blue;')
print '\nRemoving Temp Files................'
os.remove("./" + proID + curJobID + '.pdb')
os.remove("./" + proID + curJobID + '.phi')
os.remove('./param_' + curJobID)
print '\nDONE: ' + proID
except:
f = open("./" + 'ErrorLog' + curJobID + '.txt', 'w')
f.write(proID + ' error: ' + str(sys.exc_info()[0]) + str(sys.exc_info()[1]) + '\n')
f.close()
raise
def _nextSel():
sel = selection.ItemizedSelection()
sel.add(selection.currentBarrenGraphs())
if selLevel == SELRES:
items = []
addResidues = {}
for r in selection.currentResidues():
addResidues[r] = 1
selPseudoBonds = filter(
lambda e, PB=chimera.PseudoBond: isinstance(e, PB),
selection.currentEdges())
# for currently selected chain-trace pseudobonds
# act as if they select their residues
for ct in filter(
lambda e: isinstance(e.pseudoBondGroup, chimera.ChainTrace),
selPseudoBonds):
addResidues[ct.atoms[0].residue] = 1
addResidues[ct.atoms[1].residue] = 1
for r in addResidues.keys():
items.extend(r.atoms)
sel.add(items)
sel.addImplied(vertices=0)
sel.add(selPseudoBonds)
# add chain trace pseudobonds that should be selected
sel.add(selChainTrace(sel))
elif selLevel == SELCHAIN:
chainIDs = {}
for a in selection.currentAtoms():
chainID = a.residue.id.chainId
try:
chainIDs[a.molecule][chainID] = 1
except KeyError:
chainIDs[a.molecule] = {chainID: 1}
items = []
for m, ids in chainIDs.items():
for a in m.atoms:
if not ids.has_key(a.residue.id.chainId):
continue
items.append(a)
items.extend(a.bonds)
pbgs = {}
for pb in filter(lambda b, PB=chimera.PseudoBond: isinstance(b, PB),
selection.currentEdges()):
pbgs[pb.pseudoBondGroup] = 1
for pbg in pbgs.keys():
if isinstance(pbg, chimera.ChainTrace):
continue
items.extend(pbg.pseudoBonds)
sel.add(items)
sel.add(selChainTrace(sel))
elif selLevel == SELSUBMODEL:
items = []
for m in selection.currentMolecules():
items.extend(m.atoms)
items.extend(m.bonds)
items.extend(
filter(lambda e, PB=chimera.PseudoBond: isinstance(e, PB),
selection.currentEdges()))
sel.add(items)
sel.add(selChainTrace(sel))
elif selLevel == SELMODEL:
molDict = {}
for m in selection.currentMolecules():
osl = m.oslIdent()
if '.' in osl:
molDict[osl[:osl.index('.')]] = 1
else:
molDict[m] = 1
items = []
for mosl in molDict.keys():
if isinstance(mosl, basestring):
for m in selection.OSLSelection(mosl).molecules():
items.extend(m.atoms)
items.extend(m.bonds)
else:
items.extend(mosl.atoms)
items.extend(mosl.bonds)
items.extend(
filter(lambda e, PB=chimera.PseudoBond: isinstance(e, PB),
selection.currentEdges()))
sel.add(items)
sel.add(selChainTrace(sel))
elif selLevel == SELALL:
global _topValid
_topValid = True
items = []
if selection.currentMolecules():
for m in chimera.openModels.list():
if hasattr(m, 'atoms'):
items.extend(m.atoms)
items.extend(m.bonds)
for e in selection.currentEdges():
if isinstance(e, chimera.PseudoBond):
pbsSelected = True
break
else:
pbsSelected = False
if pbsSelected:
mgr = chimera.PseudoBondMgr.mgr()
for grp in mgr.pseudoBondGroups:
items.extend(grp.pseudoBonds)
sel.add(items)
sel.add(selChainTrace(sel))
global _selAllHandler
if _selAllHandler is not None:
_selAllHandler = \
chimera.openModels.addAddHandler(
_addModelHandler, None)
else:
raise ValueError, "Bad selection level"
# Handle selected surface pieces
from Surface import selected_surface_pieces
plist = selected_surface_pieces()
mlist = set([p.model for p in plist])
if selLevel == SELALL and len(mlist) > 0:
from _surface import SurfaceModel
mlist = chimera.openModels.list(modelTypes=[SurfaceModel])
sel.add(mlist)
return sel
args = cmdlineparse()
import chimera
from chimera import runCommand as rc
from Addions import initiateAddions
from DockPrep import prep
from AddCharge import estimateFormalCharge
from chimera.selection import currentResidues
if args.COMPLEX:
rc("open %s" % args.COMPLEX) # load the protein-ligand complex
if args.STRIP_IONS:
rc("delete ions")
rc("split #0 ligands")
rc("sel #0.2") # select the ligand
ligres = currentResidues()[0]
ligres.type = 'LIG' # change the resname of the ligand to 'LIG'
rc("combine #0.1 modelId 1"
) # create a new molecule containing just the receptor
rc("combine #0.2 modelId 2"
) # create a new molecule containing just the ligand
# Now that we calculated the charges of the protein and the ligand, we just need the complex
rc("combine #1,2 modelId 3"
) # create a new molecule containing the protein-ligand complex
rc("del #0-2")
pdb = args.COMPLEX.replace(".pdb", "_prep.pdb")
elif args.RECEPTOR and args.LIGAND:
rc("open %s" % args.RECEPTOR) # load the receptor
rc("open %s" % args.LIGAND) # load the ligand
if args.STRIP_IONS:
rc("delete ions")
def chimeraFeatureExtraction(
pdb_file,
radii=[5, 8, 10, 12],
metal_binding_sites=[],
disopred_disorder_map=[],
disopred_binding_map=[],
sppider_binding_map=[],
logfile="log.txt",
attempts_limit=5,
):
"""
Arguments:
metal_binding_sites: List of metal binding sites, each in dictionary
keys: 'File Name', 'Site Number', 'Metal ID', 'Residues'.
Global Features:
"""
# Open the pdb file, generate a surface, then perform DockPrep
# (Add hydrogens, etc).
# Try, in case surface generation fails.
attempts = 0
while attempts < attempts_limit:
try:
rc("open " + pdb_file)
generate_surface()
break
except Exception as e:
attempts += 1
if attempts >= attempts_limit:
raise Exception(
"Surface computation failed {} times".format(attempts))
rc("~select")
add_hydrogens_prep()
# Initalize metal binding features for this pdb file
metals = []
for metal_binding_site in metal_binding_sites:
metal_type = metal_binding_site["Metal ID"]
site_number = metal_binding_site["Site Number"]
# Convert residues from:
# C208,C211,etc ; to:
# #:208@|#:211@
residues = metal_binding_site["Residues"]
residues_selection = re.sub(r",", "@|#:", residues)
residues_selection = re.sub(r"[a-zA-Z]", "", residues_selection)
residues_selection = "#:{}@".format(residues_selection)
# Select all residues associated with this metal binding site
rc("select {}".format(residues_selection))
# Deselect backbone (only side chains should be selected)
rc("~select @n,ca,c,o")
# Clear the reply log, since we'll need it in a minute
save_and_clear_reply_log(logfile)
# Define a centroid around the currently selected residues
rc(("define centroid massWeighting false radius {} raiseTool false "
"number 1 selection").format(ionicRadiusDict[metal_type]))
# Get the x,y,z coordinates of the metal from the reply log
r, coords = read_reply_log()
coords = re.sub(r"centroid name, ID, center: centroid: c1 \( *", "",
coords)
coords = re.sub(r"\)", "", coords)
coords = re.sub(r",", "", coords)
coords = re.sub(r" +", " ", coords)
coords = re.sub(r"\n", "", coords)
coords = re.sub(r"\r", "", coords)
coords = ",".join(coords.split())
save_and_clear_reply_log(logfile)
metals.append(MetalAtom(metal_type, residues, coords, site_number))
# Depths
depths = get_depths()
# Get RPKT atoms and residues
rc("~select")
rc("select :arg@cd|:lys@ce|:mly@ce|:kcx@ce|:pro@cd|:thr@cb")
print("getting all {} rpkt residues".format(
len(selection.currentResidues())))
rpkt_residues = [
ExtendedResidue(
residue,
depths,
metals,
disopred_disorder_map,
disopred_binding_map,
sppider_binding_map,
) for residue in selection.currentResidues()
]
rpkt_atoms = [
ExtendedAtom(
atom,
depths,
metals,
disopred_disorder_map,
disopred_binding_map,
sppider_binding_map,
) for atom in selection.currentAtoms()
]
print("done")
# Get atoms and residues near RPKT
# Efficiency technique: include already-calculated RPKT atoms, delete those from selection
rc("~select")
rc("select protein")
rc("select :arg@cd z<12|:lys@ce z<12|:mly@ce z<12|:kcx@ce z<12|:pro@cd z<12|:thr@cb z<12"
)
residues_near_rpkt = list()
for residue in selection.currentResidues():
residues_near_rpkt.append(
ExtendedResidue(
residue,
depths,
metals,
disopred_disorder_map,
disopred_binding_map,
sppider_binding_map,
))
atoms_near_rpkt = list()
for atom in selection.currentAtoms():
atoms_near_rpkt.append(
ExtendedAtom(
atom,
depths,
metals,
disopred_disorder_map,
disopred_binding_map,
sppider_binding_map,
))
# Get all atoms and residues
# Efficiency technique: include already-calculated atoms near RPKT, delete those from selection
rc("~select")
rc("select protein")
rc("~select :arg@cd z<12|:lys@ce z<12|:mly@ce z<12|:kcx@ce z<12|:pro@cd z<12|:thr@cb z<12"
)
rc("~select :arg@cd|:lys@ce|:mly@ce|:kcx@ce|:pro@cd|:thr@cb")
print("getting all {} additional residues".format(
len(selection.currentResidues())))
all_residues = list(residues_near_rpkt)
for residue in selection.currentResidues():
all_residues.append(
ExtendedResidue(
residue,
depths,
metals,
disopred_disorder_map,
disopred_binding_map,
sppider_binding_map,
))
# No need to get all atoms within 12A.
print("done")
print("getting global attributes")
global_attributes = compute_global_attributes_residues(all_residues)
global_attributes.pop("circularVariance")
save_and_clear_reply_log(logfile)
print("getting bubble attributes")
atom_radius_features = {}
for radius in radii:
radius_plus_bit = radius + 1e-20
print("Analyzing at radius: {}\n".format(radius))
# Get atoms and residues near RPKT atoms
rc("~select")
rc(("select :arg@cd z<{radius}|:lys@ce z<{radius}|"
":mly@ce z<{radius}| :kcx@ce z<{radius}|:pro@cd z<{radius}|"
":thr@cb z<{radius}").format(radius=radius_plus_bit))
rc("~select :arg@cd|:lys@ce|:mly@ce|:kcx@ce|:pro@cd|:thr@cb")
print("getting all {} non-RPKT residues near rpkt".format(
len(selection.currentResidues())))
# Get all residues in the selection (near RPKT, building on previous RPKT list)
residues_near_rpkt = list(rpkt_residues)
for residue in selection.currentResidues():
residues_near_rpkt.append(
ExtendedResidue(
residue,
depths,
metals,
disopred_disorder_map,
disopred_binding_map,
sppider_binding_map,
))
atoms_near_rpkt = list(rpkt_atoms)
for atom in selection.currentAtoms():
atoms_near_rpkt.append(
ExtendedAtom(
atom,
depths,
metals,
disopred_disorder_map,
disopred_binding_map,
sppider_binding_map,
))
print("done")
# Get bubble attributes
for atom in rpkt_atoms:
if atom.name in atom_radius_features:
atom_radius_features[
atom.name][radius] = compute_bubble_attributes_residues(
atom, all_residues, atoms_near_rpkt, radius)
else:
# This must be the first radius the loop has seen.
# atom_radius_features[atom.name] should point to a dict
# with radii for keys, bubble attributes for values
atom_radius_features[atom.name] = {
radius:
compute_bubble_attributes_residues(atom, all_residues,
atoms_near_rpkt, radius)
}
print("getting atom-level attributes")
atom_features = {}
for atom in rpkt_atoms:
eResidue = ExtendedResidue(atom.residue, depths, metals)
extended_atoms = []
for a in eResidue.atoms:
extended_atoms.append(ExtendedAtom(a, depths, metals))
atom_features[atom.name] = compute_bubble_attributes_residues(
atom, residues_near_rpkt, extended_atoms, 0)
# Format each feature with _res to match labels of original pipeline data
for key in atom_features[atom.name]:
if "res" not in key:
atom_features[atom.name]["{}_res".format(key)] = atom_features[
atom.name].pop(key)
# Add global circular variance
atom_features[atom.name].update(
compute_global_circular_variance(atom, atoms_near_rpkt))
atom_features[atom.name].update({
"AA": eResidue.residue_1_letter,
"Position": eResidue.number
})
return (global_attributes, atom_features, atom_radius_features)
def Apply(self):
from chimera import selection
if not self.restrict.get() or selection.currentEmpty():
molecules = chimera.openModels.list(
modelTypes=[chimera.Molecule])
residues = []
for mol in molecules:
residues.extend(mol.residues)
else:
residues = selection.currentResidues()
molecules = tuple(set(r.molecule for r in residues))
residues = [r for r in residues
if r.ribbonResidueClass.isNucleic()]
backbone = self.showBackbone.get()
display = backbone != 'atoms & bonds'
for r in residues:
r.ribbonDisplay = display
side = self.showSide.get()
if side == 'ladder':
distSlop = 0.0
angleSlop = 0.0
relax = self.relaxParams.relaxConstraints
if relax:
distSlop = self.relaxParams.relaxDist
angleSlop = self.relaxParams.relaxAngle
NA.set_ladder(molecules, residues,
rungRadius=self.rungRadius.get(),
showStubs=self.showStubs.get(),
skipNonBaseHBonds=self.skipNonBase.get(),
useExisting=self.useExisting.get(),
distSlop=distSlop, angleSlop=angleSlop)
return
if side.endswith('slab'):
if self.currentStyle is None:
info = self._getInfo()
NA.addStyle(None, info)
showGly = self.anchor.get() != NA.SUGAR
if showGly and side.startswith('tube'):
showGly = self.showGlycosidic.get()
NA.set_slab(side, molecules, residues,
style=self.currentStyle,
thickness=self.thickness.get(),
orient=self.showOrientation.get(),
shape=self.shape.get(), showGly=showGly,
hide=self.hideBases.get())
if side.startswith('fill'):
for r in residues:
r.fillDisplay = True
else:
for r in residues:
r.fillDisplay = False
if side.endswith('fill'):
if self.showOrientation.get():
NA.set_orient(molecules, residues)
else:
NA.set_normal(molecules, residues)
elif side.startswith('atoms'):
NA.set_normal(molecules, residues)
return
def selResidues(noneReturnsAll=True, implied=False, create=False): residues = selection.currentResidues() extendSelection(residues, 'residues', noneReturnsAll, implied, create) return residues
# NOTE: the default option addHFunc=AddH.hbondAddHydrogens raised an Error in Carbonic
# Unhydrase with the Zn+2 ion. However, is works better for some proteins, where AddH.simpleAddHydrogens
# leads to net_charge prediction of the order of 120...
if args.COMPLEX:
rc("open %s" % args.COMPLEX) # load the protein-ligand complex
if args.KEEP_PROTEIN_HYDROGENS:
rc("delete element.H")
if args.REC_NET_CHARGE == None:
standardize_terminal_protein_residues(args.RECEPTOR,
"#0") # TODO: UNTESTED
if args.STRIP_IONS:
rc("delete ions")
rc("split #0 ligands")
rc("sel #0.2") # select the ligand
ligres = currentResidues()[0]
ligres.type = 'LIG' # change the resname of the ligand to 'LIG'
rc("combine #0.1 modelId 1"
) # create a new molecule containing just the receptor
rc("combine #0.2 modelId 2"
) # create a new molecule containing just the ligand
rc("del #0")
if args.REC_NET_CHARGE != None:
rec_charge = args.REC_NET_CHARGE
else:
# We will estimate the receptor's net charge. For this we need to DockPrep the receptor (is fast).
models = chimera.openModels.list(modelTypes=[chimera.Molecule])
# For a full list of DockPrep options, look into file Chimera-alpha_py2.7/share/DockPrep/__init__.py
prep([models[0]],
nogui=True,
method=args.CHARGE_METHOD,
import chimera
from Addions import initiateAddions
from DockPrep import prep
from AddCharge import estimateFormalCharge
models = chimera.openModels.list(modelTypes=[chimera.Molecule])
print(
"Preparing receptor for docking and calculating ligand AM1 charges (may be slow)."
)
prep(models, nogui=True, method='am1')
# Select the residues to be protonated
if args.RADIUS > 0:
rc("sel #1 z<%f & ~ #1" % args.RADIUS)
elif args.RADIUS == 0:
rc("sel #0")
residues = currentResidues() # get the residue of the pocket
residue_states = {}
protonatable_resids = []
protonatable_resnames = []
for r in residues:
if r.type in ["GLU", "GLH"]:
states = resname_states_dict["GLU"]
protonatable_resids.append(str(r.id))
protonatable_resnames.append(r.type)
elif r.type in ["ASP", "ASH"]:
states = resname_states_dict["ASP"]
protonatable_resids.append(str(r.id))
protonatable_resnames.append(r.type)
elif r.type in ["HIS", "HIE", "HID", "HIP"]:
states = ["HIE", "HID", "HIP"]
protonatable_resids.append(str(r.id))
def nw(s1,
s2,
scoreMatch=10,
scoreMismatch=-3,
scoreGap=0,
scoreGapOpen=-40,
gapChar=".",
returnSeqs=False,
scoreMatrix=None,
similarityMatrix=None,
frequencyMatrix=None,
endsAreGaps=False,
ssMatrix=None,
ssFraction=0.9,
gapOpenHelix=None,
gapOpenStrand=None,
gapOpenOther=None,
debug=False):
# if 'scoreMatrix', 'similarityMatrix', or 'frequencyMatrix' is
# provided, then 'scoreMatch' and 'scoreMismatch' are ignored and
# the matrix is used to evaluate matching between the sequences.
# 'scoreMatrix' should be a two-dimensional array of size
# len(s1) x len(s2). 'similarityMatrix' should be a dictionary
# keyed with two-tuples of residue types. 'frequencyMatrix' should
# be a list of length s2 of dictionaries, keyed by residue type.
#
# if 'ssFraction' is not None/False, then 'ssMatrix' should be a 3x3
# matrix keyed with 2-tuples of secondary structure types ('H': helix,
# 'S': strand, 'O': other). The score will be a mixture of the
# ss/similarity matrix scores weighted by the ssFraction
# [ssFraction * ss score + (1 - ssFraction) * similarity score]
#
# if 'gapOpenHelix/Strand/Other' is not None and 'ssFraction' is not
# None/False, then scoreGapOpen is ignored when an intra-helix/
# intra-strand/other gap is opened and the appropriate penalty
# is applied instead
#
# if 'returnSeqs' is True, then instead of returning a match list
# (a list of two-tuples) as the second value, a two-tuple of gapped
# Sequences will be returned. In both cases, the first return value
# is the match score.
m = []
bt = []
for i1 in range(len(s1) + 1):
m.append((len(s2) + 1) * [0])
bt.append((len(s2) + 1) * [None])
bt[i1][0] = 1
if endsAreGaps and i1 > 0:
m[i1][0] = scoreGapOpen + i1 * scoreGap
for i2 in range(len(s2) + 1):
bt[0][i2] = 2
if endsAreGaps and i2 > 0:
m[0][i2] = scoreGapOpen * i2 * scoreGap
if similarityMatrix is not None:
evaluate = lambda i1, i2: similarityMatrix[(s1[i1], s2[i2])]
elif scoreMatrix is not None:
evaluate = lambda i1, i2: scoreMatrix[i1][i2]
elif frequencyMatrix is not None:
evaluate = lambda i1, i2: frequencyMatrix[i2][s1[i1]]
else:
def evaluate(i1, i2):
if s1[i1] == s2[i2]:
return scoreMatch
return scoreMismatch
doingSS = ssFraction is not None and ssFraction is not False \
and ssMatrix is not None
if doingSS:
prevEval = evaluate
simFraction = 1.0 - ssFraction
def ssEval(i1, i2):
if hasattr(s1, 'ssFreqs'):
freqs1 = s1.ssFreqs[i1]
else:
freqs1 = {s1.ssType(i1): 1.0}
if hasattr(s2, 'ssFreqs'):
freqs2 = s2.ssFreqs[i2]
else:
freqs2 = {s2.ssType(i2): 1.0}
val = 0.0
for ss1, freq1 in freqs1.items():
if ss1 == None:
continue
for ss2, freq2 in freqs2.items():
if ss2 == None:
continue
val += freq1 * freq2 * ssMatrix[(ss1, ss2)]
return val
evaluate = lambda i1, i2: ssFraction * ssEval(i1, i2) + \
simFraction * prevEval(i1, i2)
# precompute appropriate gap-open penalties
gapOpen1 = [scoreGapOpen] * (len(s1) + 1)
gapOpen2 = [scoreGapOpen] * (len(s2) + 1)
if endsAreGaps:
if gapOpenOther is not None:
gapOpen1[0] = gapOpen2[0] = gapOpenOther
else:
gapOpen1[0] = gapOpen2[0] = 0
if doingSS and gapOpenOther != None:
for seq, gapOpens in [(s1, gapOpen1), (s2, gapOpen2)]:
if hasattr(seq, 'gapFreqs'):
for i, gapFreq in enumerate(seq.gapFreqs):
gapOpens[i+1] = \
gapFreq['H'] * gapOpenHelix + \
gapFreq['S'] * gapOpenStrand + \
gapFreq['O'] * gapOpenOther
else:
ssTypes = [seq.ssType(i) for i in range(len(seq))]
for i, ss in enumerate(ssTypes[:-1]):
nextSS = ssTypes[i + 1]
if ss == nextSS and ss == 'H':
gapOpens[i + 1] = gapOpenHelix
elif ss == nextSS and ss == 'S':
gapOpens[i + 1] = gapOpenStrand
else:
gapOpens[i + 1] = gapOpenOther
colGapStarts = [0] * len(s2) # don't care about column zero
for i1 in range(len(s1)):
rowGapPos = 0
for i2 in range(len(s2)):
best = m[i1][i2] + evaluate(i1, i2)
btType = 0
if i2 + 1 < len(s2) or endsAreGaps:
colGapPos = colGapStarts[i2]
skipSize = i1 + 1 - colGapPos
if hasattr(s1, "occupancy"):
totOcc = 0.0
for i in range(colGapPos, i1 + 1):
totOcc += s1.occupancy[i]
colSkipVal = totOcc * scoreGap
else:
colSkipVal = skipSize * scoreGap
baseColGapVal = m[colGapPos][i2 + 1] + colSkipVal
skip = baseColGapVal + gapOpen2[i2 + 1]
else:
skipSize = 1
colSkipVal = 0
skip = m[i1][i2 + 1]
if skip > best:
best = skip
btType = skipSize
if i1 + 1 < len(s1) or endsAreGaps:
skipSize = i2 + 1 - rowGapPos
if hasattr(s2, "occupancy"):
totOcc = 0.0
for i in range(rowGapPos, i2 + 1):
totOcc += s2.occupancy[i]
rowSkipVal = totOcc * scoreGap
else:
rowSkipVal = skipSize * scoreGap
baseRowGapVal = m[i1 + 1][rowGapPos] + rowSkipVal
skip = baseRowGapVal + gapOpen1[i1 + 1]
else:
skipSize = 1
rowSkipVal = 0
skip = m[i1 + 1][i2]
if skip > best:
best = skip
btType = 0 - skipSize
m[i1 + 1][i2 + 1] = best
bt[i1 + 1][i2 + 1] = btType
if btType >= 0:
# not gapping the row
if best > baseRowGapVal:
rowGapPos = i2 + 1
if btType <= 0:
# not gapping the column
if best > baseColGapVal:
colGapStarts[i2] = i1 + 1
if debug:
from chimera.selection import currentResidues
cr = currentResidues(asDict=True)
if cr:
for fileName, matrix in [("scores", m), ("trace", bt)]:
out = open("/home/socr/a/pett/rm/" + fileName, "w")
print >> out, " ",
for i2, r2 in enumerate(s2.residues):
if r2 not in cr:
continue
print >> out, "%5d" % i2,
print >> out
print >> out, " ",
for i2, r2 in enumerate(s2.residues):
if r2 not in cr:
continue
print >> out, "%5s" % s2[i2],
print >> out
for i1, r1 in enumerate(s1.residues):
if r1 not in cr:
continue
print >> out, "%3d" % i1, s1[i1],
for i2, r2 in enumerate(s2.residues):
if r2 not in cr:
continue
print >> out, "%5g" % (matrix[i1 + 1][i2 + 1]),
print >> out
out.close()
i1 = len(s1)
i2 = len(s2)
matchList = []
while i1 > 0 and i2 > 0:
btType = bt[i1][i2]
if btType == 0:
matchList.append((i1 - 1, i2 - 1))
i1 = i1 - 1
i2 = i2 - 1
elif btType > 0:
i1 = i1 - btType
else:
i2 = i2 + btType
if returnSeqs:
return m[len(s1)][len(s2)], matches2gappedSeqs(matchList,
s1,
s2,
gapChar=gapChar)
return m[len(s1)][len(s2)], matchList
def lists(level="atom", mode="any", attribute=None):
import chimera
from chimera import replyobj, selection
mode = findBestMatch(mode, ["any", "all"])
level = findBestMatch(level, ["atom", "residue", "chain", "molecule"])
if level == "atom":
if attribute is None:
attribute = "idatmType"
_reportAtoms(selection.currentAtoms(), attribute)
elif level == "residue":
if mode == "any":
residues = selection.currentResidues()
else:
rMap = {}
for a in selection.currentAtoms():
l = rMap.setdefault(a.residue, [])
l.append(a)
residues = []
for r, aList in rMap.iteritems():
if len(r.atoms) == len(aList):
residues.append(r)
if attribute is None:
attribute = "type"
_reportResidues(residues, attribute)
elif level == "chain":
if mode == "any":
chains = selection.currentChains()
else:
rcMap = {}
cached = set([])
cMap = {}
for r in selection.currentResidues():
if r.molecule not in cached:
cached.add(r.molecule)
for seq in r.molecule.sequences():
for res in seq.residues:
rcMap[res] = seq
try:
seq = rcMap[r]
except KeyError:
pass
else:
l = cMap.setdefault(seq, [])
l.append(r)
chains = []
for seq, rList in cMap.iteritems():
if len(seq) == len(rList):
chains.append(seq)
if attribute is None:
attribute = "chain"
_reportChains(chains, attribute)
elif level == "molecule":
if mode == "any":
molecules = selection.currentMolecules()
else:
mMap = {}
for a in selection.currentAtoms():
l = mMap.setdefault(a.molecule, [])
l.append(a)
molecules = []
for m, aList in mMap.iteritems():
if len(m.atoms) == len(aList):
molecules.append(m)
if attribute is None:
attribute = "name"
_reportModels(molecules, attribute)
else:
raise chimera.UserError("\"%s\": unknown listselection level"
% level)