def LoadTrainData4Properties(tplFile, angFile):
tpl = LoadTPL(tplFile)
Ang = LoadAngleFile(angFile)
assert tpl['sequence'] == Ang['sequence']
assert all(tpl['missing'] == Ang['Missing'])
assert (tSS == aSS
for tSS, aSS, m in zip(tpl['SS_str'], Ang['SS'], tpl['missing'])
if m != 1)
## merge tpl and ang to generate a new training and validation file
protein = dict()
protein['name'] = tpl['name']
protein['sequence'] = tpl['sequence']
protein['length'] = tpl['length']
"""
protein['NEFF'] = tpl['NEFF']
protein['PSFM'] = tpl['PSFM']
protein['PSSM'] = tpl['PSSM']
"""
protein['ACC'] = tpl['ACC']
protein['pACC'] = tpl['pACC']
protein['CNa'] = tpl['CNa']
protein['CNb'] = tpl['CNb']
#protein['Ca'] = tpl['Ca']
#protein['Cb'] = tpl['Cb']
protein['SS'] = Ang['SS']
protein['DISO'] = Ang['DISO']
protein['CLE'] = Ang['CLE']
protein['Phi'] = Ang['Phi']
protein['Psi'] = Ang['Psi']
protein['Theta'] = Ang['Theta']
protein['Tau'] = Ang['Tau']
protein['Omg'] = Ang['Omg']
##merge Phi and Psi
protein['PhiPsi'] = np.transpose(np.array([protein['Phi'],
protein['Psi']]))
##merge Theta and Tau
protein['ThetaTau'] = np.transpose(
np.array([protein['Theta'], protein['Tau']]))
## the missing residues have no 3D coordinates and thus, angles and solvent accessibility
protein['Missing'] = Ang['Missing']
protein['SS8'] = protein['SS']
protein['SS3'] = ''.join(
[PropertyUtils.SS8Letter2SS3Letter[c] for c in protein['SS8']])
protein['ForTrain'] = True
return protein
def Usage():
print 'Usage: python CheckIfSequenceHasXs.py inputFile'
print ' this script checks if the following types of files have Xs in sequence: .hhm, .hhm.pkl, .tgt, .tgt.pkl, .tpl, .tpl.pkl'
print ' if Xs are contained in the sequence, the file name will be printed out as well as the sequence'
if len(sys.argv) < 2:
Usage()
exit(1)
infile = sys.argv[1]
if infile.endswith('.pkl'):
with open(infile, 'rb') as fh:
seq = cPickle.load(fh)['sequence']
elif infile.endswith('.hhm'):
hhm = LoadHHM(infile)
seq = hhm['sequence']
elif infile.endswith('.tgt'):
tgt = LoadTGT(infile)
seq = tgt['sequence']
elif infile.endswith('.tpl'):
tpl = LoadTPL(infile)
seq = tpl['sequence']
else:
Usage()
exit(1)
if 'X' in seq:
print infile, seq
"""
if len(name) != 5:
print 'the template name is incorrect. It must be composed of PDB ID and chain letter'
exit(-1)
chain = name[4]
residues = structure[0][chain].get_residues()
residueList = [r for r in residues if is_aa(r)]
#numResidues = len(residueList)
pdbseq = ''.join([three_to_one(r.get_resname()) for r in residueList])
#print pdbseq
#from Bio import SeqIO
tpl = LoadTPL(tplfile)
tplseq = tpl['sequence']
#record = SeqIO.read(seqfile, "fasta")
##print(record.seq)
##align two sequences
from Bio import pairwise2
from Bio.SubsMat.MatrixInfo import blosum80
##alignments = pairwise2.align.localds(pdbseq, tplseq, blosum80, -5, -1)
alignments = pairwise2.align.localms(pdbseq, tplseq, 3, -1, -0.5, -0.0)
#print '#alignments:', len(alignments)
##find the alignment with the minimum residue number difference
if __name__ == "__main__":
if len(sys.argv) < 3:
Usage()
exit(1)
tplfile = sys.argv[1]
pdbfile = sys.argv[2]
ResDir = os.getcwd()
if len(sys.argv) >= 4:
ResDir = sys.argv[3]
if not os.path.isdir(ResDir):
os.mkdir(ResDir)
protein = LoadTPL(tplfile)
result, pdbseq, numMisMatches, numMatches = PDBUtils.ExtractCoordinatesNDSSPBySeq(
protein['sequence'], pdbfile)
if numMisMatches > 5:
print 'ERROR: too many mismatches between TPL sequence and ATOM record in ', pdbfile
exit(1)
if numMatches < min(30, 0.5 * len(protein['sequence'])):
print 'ERROR: more than half of TPL sequence not covered by ATOM record in ', pdbfile
exit(1)
coordInfo, dssp = result
coordinates, numInvalidAtoms = coordInfo
def ExtractCoordinatesFromTPLPDB(tplfile, pdbfile, atoms=['CB', 'CA', 'N', 'O']): tpl = LoadTPL(tplfile)