os.mkdir(savefolder)
else:
Usage()
exit(1)
if len(args) < 2:
Usage()
exit(1)
seqFile = args[0]
if not os.path.isfile(seqFile):
print "ERROR, invalid sequence file: ", seqFile
exit(1)
mainSeq = LoadFASTAFile(seqFile)
newSeq = copy.deepcopy(mainSeq)
domainModels = args[1:]
alignments = []
for dmodel in domainModels:
alignment, names = ReadAlignment(dmodel, returnNames=True)
tplSeq, tgtSeq = ExtractSeqFromAlignment(alignment)
index = newSeq.find(tgtSeq)
if index < 0:
print 'ERROR: cannot map the domain alignment sequence to the whole-chain alignment sequence'
print 'domain seq: ', tgtSeq
print 'whole seq: ', newSeq
exit(1)
## alignment[0] and alignmentp1[1] are the template and query seq in alignment, respectively
def CalcOrientationMatrix(pdbfile, sequence=None, seq_file=None):
## get pdb residues and pdb sequence
structure_id = os.path.splitext(os.path.basename(pdbfile))[0]
pdbseq, residueList = ExtractSeqFromPDBFile(pdbfile, name=structure_id)
## get sequence-pdbseq map
mapping_seq2pdb = None
if sequence is None and seq_file is None:
sequence = pdbseq
mapping_seq2pdb = np.arange(len(pdbseq))
elif sequence is None:
sequence = LoadFASTAFile(seq_file)
mapping_seq2pdb = MapSeq2PDB(sequence, pdbseq, residueList)
if mapping_seq2pdb is None:
print('ERROR: cannot map the sequence information to a pdb file: ' + pdbfile)
exit(1)
##calculate the orientation matrices
OrientationMatrix = {}
OrientationMatrix['pdbseq'] = pdbseq
OrientationMatrix['seq4matrix'] = str(sequence)
seqLen = len(sequence)
for apt in AllLabelNames:
OrientationMatrix[apt] = np.ones( (seqLen, seqLen), np.float16) * NoValidCoordinates
numInvalidAtoms=dict()
numInvalidAtoms['TwoROri']=0
numInvalidAtoms['FourCaOri']=0
for i, j in zip(range(seqLen), mapping_seq2pdb):
if j<0:
continue
residue_i = residueList[j]
for k, l in zip(range(seqLen), mapping_seq2pdb):
if l<0 :
continue
if i==k:
for apt in AllLabelNames:
OrientationMatrix[apt][i, k] = ValueOfSelf
continue
residue_k = residueList[l]
if not HasAtoms4TwoROri(residue_i, residue_k):
numInvalidAtoms['TwoROri'] += 1
pairOri = CalcOrientationOf2Residues(residue_i, residue_k)
for apt, v in pairOri.iteritems():
OrientationMatrix[apt][i, k] = v
if i == seqLen-1 or k==seqLen-1:
continue
j1 = mapping_seq2pdb[i+1]
l1 = mapping_seq2pdb[k+1]
if j1<0 or l1<0:
continue
residue_i1 = residueList[j1]
residue_k1 = residueList[l1]
if not residue_i.has_id('CA') or not residue_i1.has_id('CA') or not residue_k.has_id('CA') or not residue_k1.has_id('CA'):
numInvalidAtoms['FourCaOri'] += 1
else:
CAi = residue_i['CA']
CAi1 = residue_i1['CA']
CAk = residue_k['CA']
CAk1 = residue_k1['CA']
CaOri = CalcOrientationOf4CAs(CAi, CAi1, CAk, CAk1)
for apt, v in CaOri.iteritems():
OrientationMatrix[apt][i, k] = v
for apt in AllLabelNames:
OrientationMatrix[apt] = (OrientationMatrix[apt]).astype(np.float16)
OrientationMatrix['numInvalidAtoms'] = numInvalidAtoms
print('numInvalidAtoms: ' + str(numInvalidAtoms) )
if np.any( np.fromiter(numInvalidAtoms.itervalues(), dtype=np.int32) >5 ):
print('ERROR: there are too many invalid atoms in '+pdbfile)
return OrientationMatrix
Usage()
exit(1)
try:
opts, args = getopt.getopt(sys.argv[1:],"lc:s:",["modelList=", "cutoff=", "savefolder="])
except getopt.GetoptError:
Usage()
exit(1)
if len(args) != 2:
Usage()
exit(1)
seqFile = args[0]
target = os.path.basename(seqFile).split('.')[0]
sequence = LoadFASTAFile(seqFile)
modelListOrFolder = args[1]
if not (os.path.isdir(modelListOrFolder) or os.path.isfile(modelListOrFolder) ):
print 'ERROR: invalid model list file or folder:', modelListOrFolder
exit(1)
InputIsModelList = False
savefolder = os.getcwd()
cutoff=-1
cutoffs=[1, -1, -2]
for opt, arg in opts:
if opt in ("-l", "--modelList"):
InputIsModelList = True
else:
Usage()
exit(1)
if len(args) < 2:
Usage()
exit(1)
seqFile = args[0]
domainModels = args[1:]
if not os.path.isfile(seqFile):
print "ERROR, invalid sequence file: ", seqFile
exit(1)
mainSeq = LoadFASTAFile(seqFile)
domainSeqs = []
chainIDs = []
localQualitys = []
for dmodel in domainModels:
pdbseqs, _, chains = PDBUtils.ExtractSeqFromPDBFile(dmodel)
assert len(pdbseqs) == 1
domainSeqs.append(pdbseqs[0])
#print 'chain id: ', chains[0].get_id()
chainIDs.append(chains[0].get_id())
locQuality = ExtractLocalQuality(dmodel)
assert len(locQuality) == len(pdbseqs[0])
localQualitys.append(locQuality)
print mainSeq
def main(argv):
propertyType = 'PhiPsi'.upper()
funcType = 'AMBERPERIODIC'
allPropertyTypes = [propertyType]
allFuncTypes = ['CHARMM', 'AMBERPERIODIC', 'CIRCULARHARMONIC', 'HARMONIC']
inputFile = None
targetName = None
weight = 1.0
wStr = 'w1'
querySeqFile = None
querySeq = None
UseDisorderInfo = True
savefolder = os.getcwd()
if len(argv) < 1:
Usage()
exit(1)
try:
opts, args = getopt.getopt(argv, "a:f:w:s:q:o", [
"propertyType=", "funcType=", "weight=", "savefolder=",
"querySeqFile=", "noDisorder="
])
except getopt.GetoptError:
Usage()
exit(1)
if len(args) != 1:
Usage()
exit(1)
inputFile = args[0]
for opt, arg in opts:
if opt in ("-a", "--propertyType"):
if arg.upper() not in allPropertyTypes:
print 'ERROR: currently only support the following property types: ', allPropertyTypes
exit(1)
propertyType = arg.upper()
elif opt in ("-f", "--funcType"):
if arg.upper() not in allFuncTypes:
print 'ERROR: currently only support the following func types: ', allFuncTypes
exit(1)
funcType = arg.upper()
elif opt in ("-w", "--weight"):
weight = np.float32(arg)
wStr = 'w' + arg
if weight < 0:
print 'ERROR: the energy weight shall be >=0'
exit(1)
elif opt in ("-q", "--querySeqFile"):
querySeqFile = arg
elif opt in ("-s", "--savefolder"):
savefolder = arg
elif opt in ("-o", "--noDisorder"):
UseDisorderInfo = False
else:
Usage()
exit(1)
assert propertyType == 'PhiPsi'.upper()
if inputFile is None:
print 'ERROR: please provide an input file for predicted property'
exit(1)
if not os.path.isfile(inputFile):
print 'ERROR: the input file does not exist: ', inputFile
exit(1)
if querySeqFile is not None and os.path.isfile(querySeqFile):
querySeq = LoadFASTAFile(querySeqFile)
targetName = os.path.basename(inputFile).split('.')[0]
content = PropertyUtils.LoadPredictedProperties(inputFile)
assert len(content) >= 3
name, sequence, predProperty = content[:3]
if querySeq is not None and querySeq != sequence:
print 'ERROR: inconsistent sequences in the two files:', querySeqFile, inputFile
exit(1)
if not predProperty.has_key('PhiPsi_vonMise2d4'):
print 'ERROR: the property file does not have predicted Phi/Psi: ', inputFile
exit(1)
PhiPsiList = predProperty['PhiPsi_vonMise2d4']
predDisorder = None
if UseDisorderInfo and predProperty.has_key('disorder'):
predDisorder = predProperty['disorder']
constraints = GeneratePhiPsiPotential(sequence,
PhiPsiList,
funcType=funcType,
weight0=weight,
predDisorder=predDisorder)
if len(constraints) < 1:
print 'ERROR: cannot generate any constraints for Phi and Psi from ', inputFile
exit(1)
PhiPsiFile = os.path.join(
savefolder, targetName + '.PhiPsi4' + funcType + '.' + wStr + '.txt')
with open(PhiPsiFile, 'w') as f:
f.write('\n'.join(constraints))
def main(argv):
inputFile = None
labelNames = ['CbCb'] + config.TwoROriNames
seqSep4Dist = 1
seqSep4Ori = 2
distPotThreshold = np.finfo(np.float32).max
oriPotThreshold = 0.04
barrier = 1.0
funcType = 'SPLINE'
allFuncTypes = set(['SPLINE'])
topRatio4Ori = 25
topRatio4Dist = np.iinfo(np.int32).max
savefolder = os.getcwd()
if len(argv) < 1:
Usage()
exit(1)
try:
opts, args = getopt.getopt(argv, "a:f:s:c:b:t:q:d:", [
"labelNames=", "funcType=", "minSeqSep=", "potentialCutoff=",
"barrier=", "topRatio=", "querySeqFile=", "savefolder="
])
#print opts, args
except getopt.GetoptError:
Usage()
exit(1)
if len(args) != 1:
Usage()
exit(1)
inputFile = args[0]
querySeqFile = None
querySeq = None
for opt, arg in opts:
if opt in ("-a", "--labelNames"):
labelNames = config.ParseLabelNames(arg)
elif opt in ("-s", "--minSeqSep"):
fields = arg.split('+')
seqSep4Dist = np.int32(fields[0])
assert seqSep4Dist > 0, "The sequence separation for a valid distance potential shall be at least 1"
if len(fields) > 1:
seqSep4Ori = np.int32(fields[1])
assert seqSep4Ori > 1, "The sequence separation for a valid orientation shall be at least 2"
elif opt in ("-c", "--potentialCutoff"):
fields = arg.split('+')
oriPotThreshold = np.float32(fields[0])
if len(fields) > 1:
distPotThreshold = np.float32(fields[1])
elif opt in ("-f", "--funcType"):
funcType = arg.upper()
if funcType not in allFuncTypes:
print 'ERROR: unsupported potential func type:', funcType
exit(1)
elif opt in ("-b", "--barrier"):
barrier = np.float32(arg)
assert barrier >= 0
elif opt in ("-t", "--topRatio"):
fields = arg.split('+')
topRatio4Ori = np.float32(fields[0])
if len(fields) > 1:
topRatio4Dist = np.float32(fields[1])
elif opt in ("-q", "--querySeqFile"):
querySeqFile = arg
elif opt in ("-d", "--savefolder"):
savefolder = arg
else:
Usage()
exit(1)
if inputFile is None:
print 'ERROR: Please provide a generic distance/orientation potential file for input'
exit(1)
if not os.path.isfile(inputFile):
print 'ERROR: the input potential file does not exist: ', inputFile
exit(1)
if querySeqFile is not None and os.path.isfile(querySeqFile):
querySeq = LoadFASTAFile(querySeqFile)
if not os.path.isdir(savefolder):
os.mkdir(savefolder)
## load up the potential file
with open(inputFile, 'r') as fh:
potData = cPickle.load(fh)
if querySeq is not None and querySeq != potData[1]:
print 'ERROR: inconsistent sequences in', querySeqFile, inputFile
exit(1)
allConstraints = GenerateSplinePotential(potData,
labelNames=labelNames,
topRatio4Dist=topRatio4Dist,
topRatio4Ori=topRatio4Ori,
minSeqSep4Dist=seqSep4Dist,
minSeqSep4Ori=seqSep4Ori,
distPotThreshold=distPotThreshold,
oriPotThreshold=oriPotThreshold)
## save the Rosetta constraints into files
target = os.path.basename(inputFile).split('.')[0]
rosettaPotentialFileName = os.path.join(
savefolder, target + '.pairPotential4Rosetta.SPLINE.txt')
savefolder4histfile = os.path.join(savefolder,
'SplinePotential4' + target + '/')
WriteSplineConstraints(allConstraints,
savefile=rosettaPotentialFileName,
savefolder4histfile=savefolder4histfile)