def test_read_atom(self):
#See if we can read an atom line
line = "ATOM 41 NH1AARG A -3 12.218 84.840 88.007 0.50 40.76 N "
a = pdb_model.PdbAtom(line)
self.assertEqual(a.serial, 41)
self.assertEqual(a.name, ' NH1')
self.assertEqual(a.altLoc, 'A')
self.assertEqual(a.resName, 'ARG')
self.assertEqual(a.chainID, 'A')
self.assertEqual(a.resSeq, -3)
self.assertEqual(a.iCode, None)
self.assertEqual(a.x, 12.218)
self.assertEqual(a.y, 84.840)
self.assertEqual(a.z, 88.007)
self.assertEqual(a.occupancy, 0.5)
self.assertEqual(a.tempFactor, 40.76)
self.assertEqual(a.element, 'N')
def test_read_atom4(self):
#Round-trip an atom line
line = "ATOM 183 OD2 ASP A 24 70.534 30.495 41.026 1.00 35.00 O1-"
a = pdb_model.PdbAtom(line)
self.assertEqual(a.serial, 183)
self.assertEqual(a.name, ' OD2')
self.assertEqual(a.altLoc, None)
self.assertEqual(a.resName, 'ASP')
self.assertEqual(a.chainID, 'A')
self.assertEqual(a.resSeq, 24)
self.assertEqual(a.iCode, None)
self.assertEqual(a.x, 70.534)
self.assertEqual(a.y, 30.495)
self.assertEqual(a.z, 41.026)
self.assertEqual(a.occupancy, 1.00)
self.assertEqual(a.tempFactor, 35.00)
self.assertEqual(a.element, 'O')
self.assertEqual(a.charge, -1)
def test_read_atom3(self):
#Round-trip an atom line
line = "ATOM 160 NH1 ARG A 21 57.124 31.377 40.357 1.00 35.50 N1+"
a = pdb_model.PdbAtom(line)
self.assertEqual(a.serial, 160)
self.assertEqual(a.name, ' NH1')
self.assertEqual(a.altLoc, None)
self.assertEqual(a.resName, 'ARG')
self.assertEqual(a.chainID, 'A')
self.assertEqual(a.resSeq, 21)
self.assertEqual(a.iCode, None)
self.assertEqual(a.x, 57.124)
self.assertEqual(a.y, 31.377)
self.assertEqual(a.z, 40.357)
self.assertEqual(a.occupancy, 1.00)
self.assertEqual(a.tempFactor, 35.50)
self.assertEqual(a.element, 'N')
self.assertEqual(a.charge, 1)
def test_read_atom2(self):
#Round-trip an atom line
line = "ATOM 28 C ALA A 12 -27.804 -2.987 10.849 1.00 11.75 AA-- C "
a = pdb_model.PdbAtom(line)
self.assertEqual(a.serial, 28)
self.assertEqual(a.name, ' C ')
self.assertEqual(a.altLoc, None)
self.assertEqual(a.resName, 'ALA')
self.assertEqual(a.chainID, 'A')
self.assertEqual(a.resSeq, 12)
self.assertEqual(a.iCode, None)
self.assertEqual(a.x, -27.804)
self.assertEqual(a.y, -2.987)
self.assertEqual(a.z, 10.849)
self.assertEqual(a.occupancy, 1.00)
self.assertEqual(a.tempFactor, 11.75)
self.assertEqual(a.segID, 'AA--')
self.assertEqual(a.element, 'C')
def match_resseq(targetPdb=None, outPdb=None, resMap=None, sourcePdb=None):
assert sourcePdb or resMap
assert not (sourcePdb and resMap)
if not resMap:
resMap = residue_map.residueSequenceMap(targetPdb, sourcePdb)
chain = None # The chain we're reading
with open(targetPdb, 'r') as target, open(outPdb, 'w') as out:
for line in target:
if line.startswith("MODEL"):
raise RuntimeError("Multi-model file!")
if line.startswith("ANISOU"):
raise RuntimeError(
"I cannot cope with ANISOU! {0}".format(line))
# Stop at TER
if line.startswith("TER"):
pass
if line.startswith("ATOM"):
atom = pdb_model.PdbAtom(line)
# First atom/chain
if chain == None:
chain = atom.chainID
if atom.chainID != chain:
pass
# Get the matching resSeq for the model
modelResSeq = resMap.ref2target(atom.resSeq)
if modelResSeq == atom.resSeq:
out.write(line)
else:
atom.resSeq = modelResSeq
out.write(atom.toLine() + "\n")
continue
out.write(line)
def test_write_atom1(self):
#Round-trip an atom line
line = "ATOM 41 NH1AARG A -3 12.218 84.840 88.007 0.50 40.76 N "
a = pdb_model.PdbAtom(line)
self.assertEqual(a.toLine(), line)
def get_info(inpath):
"""Read a PDB and extract as much information as possible into a PdbInfo object
"""
info = pdb_model.PdbInfo()
info.pdb = inpath
currentModel = None
currentChain = -1
modelAtoms = [
] # list of models, each of which is a list of chains with the list of atoms
# Go through refpdb and find which ref_residues are present
f = open(inpath, 'r')
line = f.readline()
while line:
# First line of title
if line.startswith('HEADER'):
info.pdbCode = line[62:66].strip()
# First line of title
if line.startswith('TITLE') and not info.title:
info.title = line[10:-1].strip()
if line.startswith("REMARK"):
try:
numRemark = int(line[7:10])
except ValueError:
line = f.readline()
continue
# Resolution
if numRemark == 2:
line = f.readline()
if line.find("RESOLUTION") != -1:
try:
info.resolution = float(line[25:30])
except ValueError:
# RESOLUTION. NOT APPLICABLE.
info.resolution = -1
# Get solvent content
if numRemark == 280:
maxread = 5
# Clunky - read up to maxread lines to see if we can get the information we're after
# We assume the floats are at the end of the lines
for _ in range(maxread):
line = f.readline()
if line.find("SOLVENT CONTENT") != -1:
try:
info.solventContent = float(line.split()[-1])
except ValueError:
# Leave as None
pass
if line.find("MATTHEWS COEFFICIENT") != -1:
try:
info.matthewsCoefficient = float(line.split()[-1])
except ValueError:
# Leave as None
pass
# End REMARK
if line.startswith("CRYST1"):
try:
info.crystalInfo = pdb_model.CrystalInfo(line)
except ValueError as e:
logger.critical(
"ERROR READING CRYST1 LINE in file %s\":%s\"\n%s", inpath,
line.rstrip(), e)
info.crystalInfo = None
if line.startswith("MODEL"):
if currentModel:
# Need to make sure that we have an id if only 1 chain and none given
if len(currentModel.chains) <= 1:
if currentModel.chains[0] == None:
currentModel.chains[0] = 'A'
info.models.append(currentModel)
# New/first model
currentModel = pdb_model.PdbModel()
# Get serial
currentModel.serial = int(line.split()[1])
currentChain = None
modelAtoms.append([])
# Count chains (could also check against the COMPND line if present?)
if line.startswith('ATOM'):
# Create atom object
atom = pdb_model.PdbAtom(line)
# Check for the first model
if not currentModel:
# This must be the first model and there should only be one
currentModel = pdb_model.PdbModel()
modelAtoms.append([])
if atom.chainID != currentChain:
currentChain = atom.chainID
currentModel.chains.append(currentChain)
modelAtoms[-1].append([])
modelAtoms[-1][-1].append(atom)
# Can ignore TER and ENDMDL for time being as we'll pick up changing chains anyway,
# and new models get picked up by the models line
line = f.readline()
# End while loop
# End of reading loop so add the last model to the list
info.models.append(currentModel)
f.close()
bbatoms = ['N', 'CA', 'C', 'O', 'CB']
# Now process the atoms
for modelIdx, model in enumerate(info.models):
chainList = modelAtoms[modelIdx]
for chainIdx, atomList in enumerate(chainList):
# Paranoid check
assert model.chains[chainIdx] == atomList[0].chainID
# Add list of atoms to model
model.atoms.append(atomList)
# Initialise new chain
currentResSeq = atomList[0].resSeq
currentResName = atomList[0].resName
model.resSeqs.append([])
model.sequences.append("")
model.caMask.append([])
model.bbMask.append([])
atomTypes = []
for i, atom in enumerate(atomList):
aname = atom.name.strip()
if atom.resSeq != currentResSeq and i == len(atomList) - 1:
# Edge case - last residue containing one atom
atomTypes = [aname]
else:
if aname not in atomTypes:
atomTypes.append(aname)
if atom.resSeq != currentResSeq or i == len(atomList) - 1:
# End of reading the atoms for a residue
model.resSeqs[chainIdx].append(currentResSeq)
model.sequences[chainIdx] += ample_util.three2one[
currentResName]
if 'CA' not in atomTypes:
model.caMask[chainIdx].append(True)
else:
model.caMask[chainIdx].append(False)
missing = False
for bb in bbatoms:
if bb not in atomTypes:
missing = True
break
if missing:
model.bbMask[chainIdx].append(True)
else:
model.bbMask[chainIdx].append(False)
currentResSeq = atom.resSeq
currentResName = atom.resName
atomTypes = []
return info
def _keep_matching(refpdb=None, targetpdb=None, outpdb=None, resSeqMap=None):
"""Create a new pdb file that only contains that atoms in targetpdb that are
also in refpdb. It only considers ATOM lines and discards HETATM lines in the target.
Args:
refpdb: path to pdb that contains the minimal set of atoms we want to keep
targetpdb: path to the pdb that will be stripped of non-matching atoms
outpdb: output path for the stripped pdb
"""
assert refpdb and targetpdb and outpdb and resSeqMap
def _output_residue(refResidues, targetAtomList, resSeqMap, outfh):
"""Output a single residue only outputting matching atoms, shuffling the atom order and changing the resSeq num"""
# Get the matching list of atoms
targetResSeq = targetAtomList[0].resSeq
refResSeq = resSeqMap.ref2target(targetResSeq)
# Get the atomlist for the reference
for (rid, alist) in refResidues:
if rid == refResSeq:
refAtomList = alist
break
# Get ordered list of the ref atom names for this residue
rnames = [x.name for x in refAtomList]
if len(refAtomList) > len(targetAtomList):
raise RuntimeError(
"Cannot keep matching as refAtomList is > targetAtomList for residue {}\nRef: {}\nTrg: {}"
.format(targetResSeq, rnames,
[x.name for x in targetAtomList]))
# Remove any not matching in the target
alist = []
for atom in targetAtomList:
if atom.name in rnames:
alist.append(atom)
# List now only contains matching atoms
targetAtomList = alist
# Now just have matching so output in the correct order
for refname in rnames:
for i, atom in enumerate(targetAtomList):
if atom.name == refname:
# Found the matching atom
# Change resSeq and write out
atom.resSeq = refResSeq
outfh.write(atom.toLine() + "\n")
# now delete both this atom and the line
targetAtomList.pop(i)
# jump out of inner loop
break
return
# Go through refpdb and find which refResidues are present
refResidues = []
targetResSeq = [
] # ordered list of tuples - ( resSeq, [ list_of_atoms_for_that_residue ] )
last = None
chain = -1
for line in open(refpdb, 'r'):
if line.startswith("MODEL"):
raise RuntimeError("Multi-model file!")
if line.startswith("TER"):
break
if line.startswith("ATOM"):
a = pdb_model.PdbAtom(line)
# First atom/chain
if chain == -1:
chain = a.chainID
if a.chainID != chain:
raise RuntimeError(
"ENCOUNTERED ANOTHER CHAIN! {0}".format(line))
if a.resSeq != last:
last = a.resSeq
# Add the corresponding resSeq in the target
targetResSeq.append(resSeqMap.target2ref(a.resSeq))
refResidues.append((a.resSeq, [a]))
else:
refResidues[-1][1].append(a)
# Now read in target pdb and output everything bar the atoms in this file that
# don't match those in the refpdb
t = open(targetpdb, 'r')
out = open(outpdb, 'w')
chain = None # The chain we're reading
residue = None # the residue we're reading
targetAtomList = []
for line in t:
if line.startswith("MODEL"):
raise RuntimeError("Multi-model file!")
if line.startswith("ANISOU"):
raise RuntimeError("I cannot cope with ANISOU! {0}".format(line))
# Stop at TER
if line.startswith("TER"):
_output_residue(refResidues, targetAtomList, resSeqMap, out)
# we write out our own TER
out.write("TER\n")
continue
if line.startswith("ATOM"):
atom = pdb_model.PdbAtom(line)
# First atom/chain
if chain == None:
chain = atom.chainID
if atom.chainID != chain:
raise RuntimeError(
"ENCOUNTERED ANOTHER CHAIN! {0}".format(line))
if atom.resSeq in targetResSeq:
# If this is the first one add the empty tuple and reset residue
if atom.resSeq != residue:
if residue != None: # Dont' write out owt for first atom
_output_residue(refResidues, targetAtomList, resSeqMap,
out)
targetAtomList = []
residue = atom.resSeq
# If not first keep adding
targetAtomList.append(atom)
# We don't write these out as we write them with _output_residue
continue
else:
# discard this line as not a matching atom
continue
# For time being exclude all HETATM lines
elif line.startswith("HETATM"):
continue
# Endif line.startswith("ATOM")
# Output everything else
out.write(line)
# End reading loop
t.close()
out.close()
return
def read_pdb(self, pdb):
"""Get sequence as string of 1AA
get list of matching resSeq
"""
atomTypes = []
resSeq = []
resName = []
_atomTypes = []
atomTypesList = []
chain = None
readingResSeq = None
readingResName = None
for line in open(pdb):
if line.startswith("MODEL"):
raise RuntimeError("FOUND MULTI_MODEL FILE!")
if line.startswith("TER"):
break
if line.startswith("ATOM"):
atom = pdb_model.PdbAtom(line)
if not chain:
chain = atom.chainID
if atom.chainID != chain:
raise RuntimeError("FOUND ADDITIONAL CHAIN")
# First atom in first residue
if readingResSeq == None:
readingResSeq = atom.resSeq
readingResName = atom.resName
_atomTypes.append(atom.name.strip())
continue
if readingResSeq != atom.resSeq:
resName.append(readingResName)
resSeq.append(readingResSeq)
atomTypesList.append(_atomTypes)
readingResSeq = atom.resSeq
readingResName = atom.resName
_atomTypes = [atom.name.strip()]
else:
if atom.name not in _atomTypes:
_atomTypes.append(atom.name.strip())
resName.append(readingResName)
resSeq.append(readingResSeq)
atomTypesList.append(_atomTypes)
sequence = ""
for n in resName:
sequence += ample_util.three2one[n]
cAlphaMask = ['CA' not in atomTypes for atomTypes in atomTypesList]
return sequence, resSeq, cAlphaMask