def getStructFromFasta(self, fname, chainType):
'''
Creates a Bio.PDB.Structure object from a fasta file contained in fname. Atoms are not filled
and thus no coordiantes availables. Implements from Structure to Residue hierarchy.
:param fname: str. path to fasta file
@chainType: str. "l" or "r"
'''
seq = self.parseFasta(
fname, inputNumber="1" if chainType == "l" else
"2") #inpuNumber is used to report which partner fails if error
prefix = self.splitExtendedPrefix(self.getExtendedPrefix(fname))[0]
chainId = chainType.upper()
residues = []
struct = Structure(prefix)
model = Model(0)
struct.add(model)
chain = Chain(chainId)
model.add(chain)
for i, aa in enumerate(seq):
try:
resname = one_to_three(aa)
except KeyError:
resname = "UNK"
res = Residue((' ', i, ' '), resname, prefix)
chain.add(res)
return struct
def splitOnePDB(fname, outPath):
try:
s= parser.get_structure(fname, fname)
except Exception:
print ("Error loading pdb")
return 0
banLenChains=[]
try:
for chain in s[0]:
badResInChain=0
for res in chain.get_list():
if not is_aa(res,standard=True):
badResInChain+=1
chainLen= sum(1 for res in chain if "CA" in res) - badResInChain
if chainLen < MIN_SEQ_LEN or chainLen > MAX_SEQ_LEN:
print(chainLen)
banLenChains.append(chain.get_id())
except KeyError:
print ("Not good model")
return 0
for badChainId in banLenChains:
s[0].detach_child(badChainId)
receptorChainList= []
ligandChainList= []
if len( s[0].get_list())<2:
print(s)
print( s[0].get_list())
print("Not enough good chains")
return 0
for chain1 in s[0]:
tmpReceptorList=[]
for chain2 in s[0]:
if chain1!= chain2:
tmpReceptorList.append(chain2)
if len(tmpReceptorList)>1 or not tmpReceptorList[0] in ligandChainList:
ligandChainList.append(chain1)
receptorChainList.append(tmpReceptorList)
prefix= os.path.basename(fname).split(".")[0]
for i, (ligandChain, receptorChains) in enumerate(zip(ligandChainList, receptorChainList)):
io=PDBIO()
ligandStruct= Structure(prefix+"ligand")
ligandStruct.add(Model(0))
ligandChain.set_parent(ligandStruct[0])
ligandStruct[0].add(ligandChain)
io.set_structure(ligandStruct)
io.save(os.path.join(outPath,prefix+"-"+str(i)+"_l_u.pdb"))
io=PDBIO()
receptorStruct= Structure(prefix+"receptor")
receptorStruct.add(Model(0))
for receptorChain in receptorChains:
receptorChain.set_parent(receptorStruct[0])
receptorStruct[0].add(receptorChain)
io.set_structure(receptorStruct)
io.save(os.path.join(outPath,prefix+"-"+str(i)+"_r_u.pdb"))
print( "ligand:", ligandChain, "receptor:",receptorChains )
def create_new_chain(self, old_struct):
s = Structure(old_struct.chain)
my_model = Model(0)
s.add(my_model)
my_chain = Chain(old_struct.chain)
my_model.add(my_chain) #what if more chains in one component?
return s
def renumber_windowed_model(self, structure: Structure, alphafold_mmCIF_dict: Dict) -> Structure:
# Grab the Alphafold dictionary entry that descrives the residue range in the structure
seq_db_align_begin = int(alphafold_mmCIF_dict['_ma_target_ref_db_details.seq_db_align_begin'][0])
seq_db_align_end = int(alphafold_mmCIF_dict['_ma_target_ref_db_details.seq_db_align_end'][0])
# start empty
renumbered_structure = Structure(structure.id)
for model in structure:
renumbered_model = Model(model.id)
for chain in model:
transcript_residue_number = seq_db_align_begin
renumbered_chain = Chain(chain.id)
for residue in chain:
renumbered_residue = residue.copy()
renumbered_residue.id = (' ', transcript_residue_number, ' ')
# The above copy routines fail to copy disorder properly - so just wipe out all notion of disorder
for atom in renumbered_residue:
atom.disordered_flag = 0
renumbered_residue.disordered = 0
renumbered_chain.add(renumbered_residue)
transcript_residue_number += 1
assert transcript_residue_number == seq_db_align_end + 1
renumbered_model.add(renumbered_chain)
renumbered_structure.add(renumbered_model)
return renumbered_structure
def create_new_chain(self, id):
"""
"""
self.fragment_lattice = Structure(id)
my_model = Model(0)
self.fragment_lattice.add(my_model)
my_chain = Chain(id)
my_model.add(my_chain) #what if more chains in one component?
def initialize_res(residue: Union[Geo, str]) -> Structure:
"""Creates a new structure containing a single amino acid. The type and
geometry of the amino acid are determined by the argument, which has to be
either a geometry object or a single-letter amino acid code.
The amino acid will be placed into chain A of model 0."""
if isinstance(residue, Geo):
geo = residue
elif isinstance(residue, str):
geo = geometry(residue)
else:
raise ValueError("Invalid residue argument:", residue)
segID = 1
AA = geo.residue_name
CA_N_length = geo.CA_N_length
CA_C_length = geo.CA_C_length
N_CA_C_angle = geo.N_CA_C_angle
CA_coord = np.array([0.0, 0.0, 0.0])
C_coord = np.array([CA_C_length, 0, 0])
N_coord = np.array([
CA_N_length * math.cos(N_CA_C_angle * (math.pi / 180.0)),
CA_N_length * math.sin(N_CA_C_angle * (math.pi / 180.0)),
0,
])
N = Atom("N", N_coord, 0.0, 1.0, " ", " N", 0, "N")
# Check if the peptide is capped or not
if geo.residue_name == "ACE":
CA = Atom("CH3", CA_coord, 0.0, 1.0, " ", " CH3", 0, "C")
else:
CA = Atom("CA", CA_coord, 0.0, 1.0, " ", " CA", 0, "C")
C = Atom("C", C_coord, 0.0, 1.0, " ", " C", 0, "C")
##Create Carbonyl atom (to be moved later)
C_O_length = geo.C_O_length
CA_C_O_angle = geo.CA_C_O_angle
N_CA_C_O_diangle = geo.N_CA_C_O_diangle
carbonyl = calculateCoordinates(N, CA, C, C_O_length, CA_C_O_angle,
N_CA_C_O_diangle)
O = Atom("O", carbonyl, 0.0, 1.0, " ", " O", 0, "O")
res = make_res_of_type(segID, N, CA, C, O, geo)
cha = Chain("A")
cha.add(res)
mod = Model(0)
mod.add(cha)
struc = Structure("X")
struc.add(mod)
return struc
def init_model(self, model_id, serial_num=None):
"""Initiate a new Model object with given id.
Arguments:
o id - int
o serial_num - int
"""
self.model = Model(model_id, serial_num)
self.structure.add(self.model)
def splitOnePDB(fname, chainIdL, chainIdR, outPath):
print(os.path.basename(fname))
try:
s = parser.get_structure(os.path.basename(fname), fname)
except Exception:
print("Error loading pdb")
return 0
banLenChains = []
try:
for chain in s[0]:
badResInChain = 0
for res in chain.get_list():
if not is_aa(res, standard=True) and res.resname != "HOH":
badResInChain += 1
# for res in chain: print(res)
chainLen = sum(1 for res in chain if "CA" in res) - badResInChain
if chainLen < MIN_SEQ_LEN or chainLen > MAX_SEQ_LEN:
print(chain, chainLen)
banLenChains.append(chain.get_id())
except KeyError:
print("Not good model")
return 0
# print(banLenChains)
if len(s[0].get_list()) - len(banLenChains) < 2:
print(s)
print(s[0].get_list())
print("Not enough good chains")
return 0
ligandChains, receptorChains = findNeigChains(s, chainIdL, chainIdR)
print("ligand:", ligandChains, "receptor:", receptorChains)
prefix = os.path.basename(fname).split(".")[0]
io = PDBIO()
ligandStruct = Structure(prefix + "ligand")
ligandStruct.add(Model(0))
for ligandChain in ligandChains:
ligandChain.set_parent(ligandStruct[0])
ligandStruct[0].add(ligandChain)
io.set_structure(ligandStruct)
io.save(
os.path.join(outPath, prefix + "-" + chainIdL + chainIdR + "_l_u.pdb"))
io = PDBIO()
receptorStruct = Structure(prefix + "receptor")
receptorStruct.add(Model(0))
for receptorChain in receptorChains:
receptorChain.set_parent(receptorStruct[0])
receptorStruct[0].add(receptorChain)
io.set_structure(receptorStruct)
io.save(
os.path.join(outPath, prefix + "-" + chainIdL + chainIdR + "_r_u.pdb"))
def single_chain_structure(chain, name='superposition'):
from Bio.PDB.Structure import Structure
from Bio.PDB.Model import Model
structure = Structure(name)
model = Model(0)
structure.add(model)
model.add(chain)
return structure
def complex_save(given_complex, i, path):
s = Structure(i)
my_model = Model(0)
s.add(my_model)
for component in given_complex.components:
my_model.add(
component.pyrystruct.struct[0][component.pyrystruct.chain])
out = PDBIO()
out.set_structure(s)
out.save(path)
return path
def slice(cls, obj, selection, name='slice'):
"""Create a new Structure object 'S2' from a slice of the current one, 'S1'. <selection>
defines which descendents 'S1' will be stored in 'S2'."""
from Bio.PDB.Structure import Structure
from Bio.PDB.Model import Model
from Bio.PDB.Chain import Chain
ent = Structure(name) # Biopython structure object
# Loop over selection and determine what model/chain objects we need to create in order to
# store the slice
models = {}
for item in selection:
mid = item[1]
cid = item[2]
if mid not in models:
models[mid] = set() # store chain ids
models[mid].add(cid)
# Create model/chains to store slice
for mid in models:
ent.add(Model(mid))
for cid in models[mid]:
ent[mid].add(Chain(cid))
# Add residues to slice
for item in selection:
mid = item[1]
cid = item[2]
rid = item[3]
ent[mid][cid].add(obj[mid][cid][rid].copy())
return cls(ent, name=name)
def add(self, residue):
"""Add PdbResidue object to site (in the residues list and dict)"""
residue = residue.copy(include_structure=True)
if type(residue) == PdbResidue:
self.residues.append(residue)
self.residues_dict[residue.full_id] = residue
residue.parent_site = self
if type(residue) == Het:
self.ligands.append(residue)
residue.parent_site = self
if residue.is_polymer:
if residue.chain in self.structure[0]:
for r in residue.structure:
self.structure[0][residue.chain].add(r)
return True
self.structure[0].add(residue.structure)
return True
if residue.structure:
# Initialize structure if empty
if self.structure is None:
self.structure = Structure(self.id)
self.structure.add(Model(0))
chain_id = residue.structure.get_parent().get_id()
if chain_id not in self.structure[0]:
self.structure[0].add(Chain(chain_id))
# Add residue structure to site structure
if residue.structure.get_id() not in self.structure[0][chain_id]:
self.structure[0][chain_id].add(residue.structure)
return True
def retrieve_sphere_model(structure): #, score):
"""
each chain is here represented by centre of mass only
"""
sphere_struct = Structure('clustering_model')
my_model = Model(0)
sphere_struct.add(my_model)
#bedzie zmieniona numeracja
chain_mass_centres, index = [], 0
for chain in structure.get_chains():
my_chain = Chain(chain.id)
sphere_struct[0].add(my_chain)
coord = calculate_centre_of_complex(chain)
chain_mass_centres.append(coord)
my_residue = Residue((' ', index, ' '), chain.id, ' ')
coords = array(coord, 'f')
atom = Atom('CA', coords, 0, 0, ' ', 'CA', 1)
my_chain.add(my_residue)
my_residue.add(atom)
index += 1
del structure
return sphere_struct
def retrieve_ca_model(structure):
"""
chains are represented only by main chain atoms (Calfas or C4')
"""
reduced_struct = Structure('clustering_model')
my_model = Model(0)
reduced_struct.add(my_model)
main_chain_atoms = []
for ch in structure[0]:
my_chain = Chain(ch.id)
reduced_struct[0].add(my_chain)
for resi in ch:
for atom in resi:
#print "----", resi.id, resi.get_segid(), ch.id
if atom.get_name() == "CA" or atom.get_name(
) == "C4'" or atom.get_name() == "C4*":
my_residue = Residue((' ', resi.id[1], ' '),
resi.get_resname(), ' ')
atom = Atom('CA', atom.coord, 0, ' ', ' ', 'CA',
atom.get_serial_number())
my_chain.add(my_residue)
my_residue.add(atom)
main_chain_atoms.append(atom)
return reduced_struct
def create_sphere_representation(self):
"""
each chain is here represented by centre of mass only
"""
new_struct = Structure('sphrere')
my_model = Model(0)
new_struct.add(my_model)
chain_mass_centres, index = [], 1
my_chain = Chain(self.fa_struct.chain)
new_struct[0].add(my_chain)
coord, self.molmass, self.radius = self.calculate_centre_of_complex(
self.fa_struct.struct)
my_residue = Residue((' ', index, ' '), "ALA", ' ')
coords = array(coord, 'f')
atom = Atom('CA', coords, 0, 0, ' ', ' CA', 1)
my_chain.add(my_residue)
my_residue.add(atom)
self.cg_struct = new_struct
name = "dddd" + self.fa_struct.chain
self.save_pdb(new_struct, name)
def __create_superimposed_pdb(self):
def fill_in_chain(chain, protein_id, rotation_matrix = None):
for index,residue in enumerate(self.proteins[protein_id].get_residues()):
residue.id = (residue.id[0], index, residue.id[2])
chain.add(residue)
merged_model = Model(0)
chain_a = Chain('A')
chain_b = Chain('B')
fill_in_chain(chain_a, 0)
fill_in_chain(chain_b, 1)
merged_model.add(chain_a)
merged_model.add(chain_b)
return merged_model
def init_model(self, model_id, serial_num = None):
"""Initiate a new Model object with given id.
Arguments:
o id - int
o serial_num - int
"""
self.model=Model(model_id,serial_num)
self.structure.add(self.model)
def initialize_res(residue):
'''Creates a new structure containing a single amino acid. The type and
geometry of the amino acid are determined by the argument, which has to be
either a geometry object or a single-letter amino acid code.
The amino acid will be placed into chain A of model 0.'''
if isinstance( residue, Geo ):
geo = residue
else:
geo= Geo(residue)
segID=1
AA= geo.residue_name
CA_N_length=geo.CA_N_length
CA_C_length=geo.CA_C_length
N_CA_C_angle=geo.N_CA_C_angle
CA_coord= np.array([0.,0.,0.])
C_coord= np.array([CA_C_length,0,0])
N_coord = np.array([CA_N_length*math.cos(N_CA_C_angle*(math.pi/180.0)),CA_N_length*math.sin(N_CA_C_angle*(math.pi/180.0)),0])
N= Atom("N", N_coord, 0.0 , 1.0, " "," N", 0, "N")
CA=Atom("CA", CA_coord, 0.0 , 1.0, " "," CA", 0,"C")
C= Atom("C", C_coord, 0.0, 1.0, " ", " C",0,"C")
##Create Carbonyl atom (to be moved later)
C_O_length=geo.C_O_length
CA_C_O_angle=geo.CA_C_O_angle
N_CA_C_O_diangle=geo.N_CA_C_O_diangle
carbonyl=calculateCoordinates(N, CA, C, C_O_length, CA_C_O_angle, N_CA_C_O_diangle)
O= Atom("O",carbonyl , 0.0 , 1.0, " "," O", 0, "O")
res=makeRes(segID, N, CA, C, O, geo)
cha= Chain('A')
cha.add(res)
mod= Model(0)
mod.add(cha)
struc= Structure('X')
struc.add(mod)
return struc
def save_pdb(self, complex_id, temp = "", name = ""):
"""
gets coordinates of all complex components and writes them in one
file one component = one pdb model
Parameters:
------------
complex_id : number of complex from simulation
Returns:
--------
pdb files with simulated components in OUTFOLDER
"""
##add component chain by chain not residue by residue.
model_num = 0
score = round(self.simulation_score, 4)
s = Structure(complex_id)
my_model = Model(0)
s.add(my_model)
for component in self.components:
#@TODO: #what if more chains in one component?
my_model.add(component.pyrystruct.struct[0][component.pyrystruct.chain])
out = PDBIO()
out.set_structure(s)
outname = outfolder.outdirname.split("/")[-1]
temp = str(temp)
try:
temp = round(float(temp),1)
except: pass
if name:
fi_name = str(outfolder.outdirname)+'/'+name+'_'+str(score)+'_'+str(complex_id)+"_"+str(temp)+'.pdb'
out.save(fi_name)
else:
fi_name = str(outfolder.outdirname)+'/'+str(outname)+"_"+str(score)+'_'+str(complex_id)+"_"+str(temp)+'.pdb'
out.save(fi_name)
for comp in self.components:
comp.pyrystruct.struct[0][comp.pyrystruct.chain].detach_parent()
return fi_name
def get_structure(self, name='RNA chain'):
"""Returns chain as a PDB.Structure object."""
struc = Structure(name)
model = Model(0)
chain = Chain(self.chain_name)
struc.add(model)
struc[0].add(chain)
for resi in self:
struc[0][self.chain_name].add(resi)
return struc
def calculate_BSA(self):
"Uses NACCESS module in order to calculate the Buried Surface Area"
# Extract list of chains in the interface only
chains = list(self.get_chains())
# Create temporary structures to feed NACCESS
structure_A=Structure("chainA")
structure_B=Structure("chainB")
mA = Model(0)
mB = Model(0)
mA.add(self.model[chains[0]])
mB.add(self.model[chains[1]])
structure_A.add(mA)
structure_B.add(mB)
# Calculate SASAs
NACCESS_atomic(self.model)
NACCESS_atomic(structure_A[0])
NACCESS_atomic(structure_B[0])
sas_tot= _get_atomic_SASA(self.model)
#print 'Accessible surface area, complex:', sas_tot
sas_A= _get_atomic_SASA(structure_A)
#print 'Accessible surface aream CHAIN A :', sas_A
sas_B= _get_atomic_SASA(structure_B)
#print 'Accessible surface aream CHAIN B :',sas_B
# Calculate BSA
bsa = sas_A+sas_B-sas_tot
return [bsa, sas_A, sas_B, sas_tot]
def select_structure(selector, structure):
new_structure = Structure(structure.id)
for model in structure:
if not selector.accept_model(model):
continue
new_model = Model(model.id, model.serial_num)
new_structure.add(new_model)
for chain in model:
if not selector.accept_chain(chain):
continue
new_chain = Chain(chain.id)
new_model.add(new_chain)
for residue in chain:
if not selector.accept_residue(residue):
continue
new_residue = Residue(residue.id, residue.resname,
residue.segid)
new_chain.add(new_residue)
for atom in residue:
if selector.accept_atom(atom):
new_residue.add(atom)
return new_structure
def create_structure(coords, pdb_type, remove_masked):
"""Create the structure.
Args:
coords: 3D coordinates of structure
pdb_type: predict or actual structure
remove_masked: whether to include masked atoms. If false,
the masked atoms have coordinates of [0,0,0].
Returns:
structure
"""
name = protein.id_
structure = Structure(name)
model = Model(0)
chain = Chain('A')
for i, residue in enumerate(protein.primary):
residue = AA_LETTERS[residue]
if int(protein.mask[i]) == 1 or remove_masked == False:
new_residue = Residue((' ', i + 1, ' '), residue, ' ')
j = 3 * i
atom_list = ['N', 'CA', 'CB']
for k, atom in enumerate(atom_list):
new_atom = Atom(name=atom,
coord=coords[j + k, :],
bfactor=0,
occupancy=1,
altloc=' ',
fullname=" {} ".format(atom),
serial_number=0)
new_residue.add(new_atom)
chain.add(new_residue)
model.add(chain)
structure.add(model)
io = PDBIO()
io.set_structure(structure)
io.save(save_dir + name + '_' + pdb_type + '.pdb')
return structure
def _rsa_calculation(self, model, chain_list, rsa_threshold):
"Uses NACCESS module in order to calculate the Buried Surface Area"
pairs=[]
# Create temporary structures to feed NACCESS
structure_A=Structure("chainA")
structure_B=Structure("chainB")
mA = Model(0)
mB = Model(0)
mA.add(model[chain_list[0]])
mB.add(model[chain_list[1]])
structure_A.add(mA)
structure_B.add(mB)
# Calculate SASAs
nacc_at=NACCESS(model)
model_values=[]
res_list = [r for r in model.get_residues() if r.id[0] == ' ']
structure_A_reslist =[r for r in structure_A[0].get_residues() if r.id[0] == ' ']
structure_B_reslist =[r for r in structure_B[0].get_residues() if r.id[0] == ' ']
for res in res_list:
model_values.append(float(res.xtra['EXP_NACCESS']['all_atoms_rel']))
sas_tot= self._get_residue_SASA(model)
#print 'Accessible surface area, complex:', sas_tot
nacc_at=NACCESS(structure_A[0])
nacc_at=NACCESS(structure_B[0])
submodel_values=[]
for res in structure_A_reslist:
if res.id[0]==' ':
submodel_values.append(float(res.xtra['EXP_NACCESS']['all_atoms_rel']))
for res in structure_B_reslist:
if res.id[0]==' ':
submodel_values.append(float(res.xtra['EXP_NACCESS']['all_atoms_rel']))
count=0
for res in res_list:
if res in structure_A_reslist and ((submodel_values[count] - model_values[count]) > rsa_threshold):
pairs.append(res)
elif res in structure_B_reslist and ((submodel_values[count] - model_values[count]) > rsa_threshold):
pairs.append(res)
count=count+1
sas_A= self._get_residue_SASA(structure_A)
#print 'Accessible surface aream CHAIN A :', sas_A
sas_B= self._get_residue_SASA(structure_B)
#print 'Accessible surface aream CHAIN B :',sas_B
# Calculate BSA
bsa = sas_A+sas_B-sas_tot
self.interface.accessibility=[bsa, sas_A, sas_B, sas_tot]
return pairs
def createPDBFile(self):
"Create test CIF file with 12 Atoms in icosahedron vertexes"
from Bio.PDB.Structure import Structure
from Bio.PDB.Model import Model
from Bio.PDB.Chain import Chain
from Bio.PDB.Residue import Residue
from Bio.PDB.Atom import Atom
from Bio.PDB.mmcifio import MMCIFIO
import os
CIFFILENAME = "/tmp/out.cif"
# create atom struct with ico simmety (i222r)
icosahedron = Icosahedron(circumscribed_radius=100, orientation='222r')
pentomVectorI222r = icosahedron.getVertices()
# create biopython object
structure = Structure('result') # structure_id
model = Model(1, 1) # model_id,serial_num
structure.add(model)
chain = Chain('A') # chain Id
model.add(chain)
for i, v in enumerate(pentomVectorI222r, 1):
res_id = (' ', i, ' ') # first arg ' ' -> aTOm else heteroatom
res_name = "ALA" #+ str(i) # define name of residue
res_segid = ' '
residue = Residue(res_id, res_name, res_segid)
chain.add(residue)
# ATOM name, coord, bfactor, occupancy, altloc, fullname, serial_number,
# element=None)
atom = Atom('CA', v, 0., 1., " ", " CA ", i, "C")
residue.add(atom)
io = MMCIFIO()
io.set_structure(structure)
# delete file if exists
if os.path.exists(CIFFILENAME):
os.remove(CIFFILENAME)
io.save(CIFFILENAME)
return CIFFILENAME
def from_domain(cls,
domain: DomainResidueMapping,
bio_structure: Model,
residue_id_mapping: BiopythonToMmcifResidueIds.Mapping,
skip_label_seq_id=lambda id: False):
bio_chain = bio_structure[
domain.chain_id] # todo wtf proč??? proč tam nepošlu rovnou chain?
domain_residues = [
bio_chain[residue_id_mapping.to_bio(label_seq_id)]
for label_seq_id in domain if not skip_label_seq_id(label_seq_id)
]
return cls(domain_residues,
bio_structure.get_parent().id, domain.chain_id,
domain.domain_id)
def create_new_structure(self, name, chain_id):
"""
creates new Bio.PDB structure object
Parameters:
-----------
name : structure name
chain_id : chain name (e.g. A, B, C)
Returns:
---------
self.struct : Bio.PDB object with model and chain inside
"""
self.struct = Structure(name)
my_model = Model(0)
my_chain = Chain(chain_id)
self.struct.add(my_model)
self.struct[0].add(my_chain)
def superimpose_structure(s1: Model, by_1: SetOfResidues, on_2: SetOfResidues):
""" Rotate and move the whole s1 structure on a second structure. So that by_1 will be superimposed onto on_2.
by_1 and on_2 can be a subset of the structure's residues (e.g. domains). Both have to have the same length and
should contain corresponding (sequence aligned) residues, in the same order.
:returns a copy of s1 Model with all its atom coordinates rotated and translated (by the superposition).
"""
s1_copy = s1.copy()
# put all coordinates of s1 into a np.array
# (get_unpacked_list [in contrary to Model.get_atoms()] also includes disordered atoms/residues. So this would
# handle that correctly, however we skip structures with disordered residues anyway in our analyses.)
s1_atoms = [a for c in s1_copy.get_chains() for r in c.get_unpacked_list() for a in r.get_unpacked_list()]
s1_coords = np.array([atom.get_coord() for atom in s1_atoms])
# superimpose the two structures by their first domain
U = get_rotation_matrix(by_1, on_2)
new_s1_coords = (s1_coords - get_centroid(by_1)) @ U + get_centroid(on_2)
for atom, new_coord in zip(s1_atoms, new_s1_coords):
atom.set_coord(new_coord)
return s1_copy
def __make_structure_from_residues__(self, residues):
"""
Makes a Structure object either from a pdbfile or a list of residues
"""
# KR: this probably can be outsourced to another module.
struct = Structure('s')
model = Model('m')
n_chain = 1
chain = Chain('c%i' % n_chain)
for residue in residues:
if chain.has_id(residue.id):
model.add(chain)
n_chain += 1
chain = Chain('c%i' % n_chain)
chain.add(residue)
model.add(chain)
struct.add(model)
return struct
class StructureBuilder(object):
"""
Deals with contructing the Structure object. The StructureBuilder class is used
by the PDBParser classes to translate a file to a Structure object.
"""
def __init__(self):
self.line_counter=0
self.header={}
def _is_completely_disordered(self, residue):
"Return 1 if all atoms in the residue have a non blank altloc."
atom_list=residue.get_unpacked_list()
for atom in atom_list:
altloc=atom.get_altloc()
if altloc==" ":
return 0
return 1
# Public methods called by the Parser classes
def set_header(self, header):
self.header=header
def set_line_counter(self, line_counter):
"""
The line counter keeps track of the line in the PDB file that
is being parsed.
Arguments:
o line_counter - int
"""
self.line_counter=line_counter
def init_structure(self, structure_id):
"""Initiate a new Structure object with given id.
Arguments:
o id - string
"""
self.structure=Structure(structure_id)
def init_model(self, model_id, serial_num = None):
"""Initiate a new Model object with given id.
Arguments:
o id - int
o serial_num - int
"""
self.model=Model(model_id,serial_num)
self.structure.add(self.model)
def init_chain(self, chain_id):
"""Initiate a new Chain object with given id.
Arguments:
o chain_id - string
"""
if self.model.has_id(chain_id):
self.chain=self.model[chain_id]
warnings.warn("WARNING: Chain %s is discontinuous at line %i."
% (chain_id, self.line_counter),
PDBConstructionWarning)
else:
self.chain=Chain(chain_id)
self.model.add(self.chain)
def init_seg(self, segid):
"""Flag a change in segid.
Arguments:
o segid - string
"""
self.segid=segid
def init_residue(self, resname, field, resseq, icode):
"""
Initiate a new Residue object.
Arguments:
o resname - string, e.g. "ASN"
o field - hetero flag, "W" for waters, "H" for
hetero residues, otherwise blank.
o resseq - int, sequence identifier
o icode - string, insertion code
"""
if field!=" ":
if field=="H":
# The hetero field consists of H_ + the residue name (e.g. H_FUC)
field="H_"+resname
res_id=(field, resseq, icode)
if field==" ":
if self.chain.has_id(res_id):
# There already is a residue with the id (field, resseq, icode).
# This only makes sense in the case of a point mutation.
warnings.warn("WARNING: Residue ('%s', %i, '%s') "
"redefined at line %i."
% (field, resseq, icode, self.line_counter),
PDBConstructionWarning)
duplicate_residue=self.chain[res_id]
if duplicate_residue.is_disordered()==2:
# The residue in the chain is a DisorderedResidue object.
# So just add the last Residue object.
if duplicate_residue.disordered_has_id(resname):
# The residue was already made
self.residue=duplicate_residue
duplicate_residue.disordered_select(resname)
else:
# Make a new residue and add it to the already
# present DisorderedResidue
new_residue=Residue(res_id, resname, self.segid)
duplicate_residue.disordered_add(new_residue)
self.residue=duplicate_residue
return
else:
# Make a new DisorderedResidue object and put all
# the Residue objects with the id (field, resseq, icode) in it.
# These residues each should have non-blank altlocs for all their atoms.
# If not, the PDB file probably contains an error.
if not self._is_completely_disordered(duplicate_residue):
# if this exception is ignored, a residue will be missing
self.residue=None
raise PDBConstructionException(
"Blank altlocs in duplicate residue %s ('%s', %i, '%s')"
% (resname, field, resseq, icode))
self.chain.detach_child(res_id)
new_residue=Residue(res_id, resname, self.segid)
disordered_residue=DisorderedResidue(res_id)
self.chain.add(disordered_residue)
disordered_residue.disordered_add(duplicate_residue)
disordered_residue.disordered_add(new_residue)
self.residue=disordered_residue
return
residue=Residue(res_id, resname, self.segid)
self.chain.add(residue)
self.residue=residue
def init_atom(self, name, coord, b_factor, occupancy, altloc, fullname,
serial_number=None, element=None):
"""
Initiate a new Atom object.
Arguments:
o name - string, atom name, e.g. CA, spaces should be stripped
o coord - Numeric array (Float0, size 3), atomic coordinates
o b_factor - float, B factor
o occupancy - float
o altloc - string, alternative location specifier
o fullname - string, atom name including spaces, e.g. " CA "
o element - string, upper case, e.g. "HG" for mercury
"""
residue=self.residue
# if residue is None, an exception was generated during
# the construction of the residue
if residue is None:
return
# First check if this atom is already present in the residue.
# If it is, it might be due to the fact that the two atoms have atom
# names that differ only in spaces (e.g. "CA.." and ".CA.",
# where the dots are spaces). If that is so, use all spaces
# in the atom name of the current atom.
if residue.has_id(name):
duplicate_atom=residue[name]
# atom name with spaces of duplicate atom
duplicate_fullname=duplicate_atom.get_fullname()
if duplicate_fullname!=fullname:
# name of current atom now includes spaces
name=fullname
warnings.warn("Atom names %r and %r differ "
"only in spaces at line %i."
% (duplicate_fullname, fullname,
self.line_counter),
PDBConstructionWarning)
atom=self.atom=Atom(name, coord, b_factor, occupancy, altloc,
fullname, serial_number, element)
if altloc!=" ":
# The atom is disordered
if residue.has_id(name):
# Residue already contains this atom
duplicate_atom=residue[name]
if duplicate_atom.is_disordered()==2:
duplicate_atom.disordered_add(atom)
else:
# This is an error in the PDB file:
# a disordered atom is found with a blank altloc
# Detach the duplicate atom, and put it in a
# DisorderedAtom object together with the current
# atom.
residue.detach_child(name)
disordered_atom=DisorderedAtom(name)
residue.add(disordered_atom)
disordered_atom.disordered_add(atom)
disordered_atom.disordered_add(duplicate_atom)
residue.flag_disordered()
warnings.warn("WARNING: disordered atom found "
"with blank altloc before line %i.\n"
% self.line_counter,
PDBConstructionWarning)
else:
# The residue does not contain this disordered atom
# so we create a new one.
disordered_atom=DisorderedAtom(name)
residue.add(disordered_atom)
# Add the real atom to the disordered atom, and the
# disordered atom to the residue
disordered_atom.disordered_add(atom)
residue.flag_disordered()
else:
# The atom is not disordered
residue.add(atom)
def set_anisou(self, anisou_array):
"Set anisotropic B factor of current Atom."
self.atom.set_anisou(anisou_array)
def set_siguij(self, siguij_array):
"Set standard deviation of anisotropic B factor of current Atom."
self.atom.set_siguij(siguij_array)
def set_sigatm(self, sigatm_array):
"Set standard deviation of atom position of current Atom."
self.atom.set_sigatm(sigatm_array)
def get_structure(self):
"Return the structure."
# first sort everything
# self.structure.sort()
# Add the header dict
self.structure.header=self.header
return self.structure
def set_symmetry(self, spacegroup, cell):
pass
""" Read uniformly distributed sphere points from file
"""
num_lines = sum(1 for line in open(filename))
points = zeros(shape=(num_lines,3))
ind = 0
for line in open(filename).readlines():
points[ind] = array((line.split()[0:3]))
points[ind] = points[ind] * scale
ind = ind+1
return points
#--------------------------------------------------------------------
# ref_ptsfilename = "K562.pts"
refid = "ref"
structure = Structure(refid)
model_ref = Model(1)
chain_ref = Chain("A")
points_ref = ReadXYZ(ref_ptsfilename,scale)
num_count = 0
for i in range(0,shape(points_ref[IndexList])[0]):
num_count = num_count +1
# res_id = (' ',IndexList[i],' ')
res_id = (' ',num_count,' ')
residue = Residue(res_id,'ALA',' ')
cur_coord = tuple(points_ref[IndexList[i]])
# atom = Atom('CA',cur_coord,0,0,' ',num_count,num_count,'C')
atom = Atom('CA',cur_coord,0,0,' ',num_count,num_count,'C')
residue.add(atom)
chain_ref.add(residue)
model_ref.add(chain_ref)
all_models_stats = []
files = [f for f in os.listdir(path) if os.path.isfile(join(path,f)) and '.pdb' in f]
files.sort()
for pdb_file in files:#os.listdir(path):
print pdb_file + '\n'
parser = PDBParser()
struct = parser.get_structure('structure', path + pdb_file)
chains = list(struct[0].get_chains())
#print str(len(chains))
compares = []#storing compares that are already done
#analyzed_count += 1
model = Model()
model.name = pdb_file
for ch1 in range(0,len(chains)):
for ch2 in range(ch1 + 1,len(chains)):
checklist = [ chains[ch1].get_full_id()[2] , chains[ch2].get_full_id()[2] ]
checklist2 = [ chains[ch2].get_full_id()[2], chains[ch1].get_full_id()[2] ]
if chains[ch1].get_full_id()[2] != chains[ch2].get_full_id()[2] and not checklist in compares and not checklist2 in compares:
comparsion = [ chains[ch1].get_full_id()[2] , chains[ch2].get_full_id()[2] ]
compares.append(comparsion)#appending comprasion to already done comprasions
chain1_atms = list(chains[ch1].get_atoms())
chain2_atms = list(chains[ch2].get_atoms())
#print str(len(chain1_atms)) + ' ' + str(len(chain2_atms))
bfactors = zeros(shape=(num_lines,1))
ind = 0
for line in open(filename).readlines():
if not line.startswith('#'):
bfactors[ind] = array((line.split())[column])
ind = ind+1
return bfactors
#--------------------------------------------------------------------
points = ReadXYZ ( args['src'], args['scale'])
if ( args['bfactor'] is not None):
print "read bfactor file column %d" % args['column']
bfactors = ReadBfactor(args['bfactor'],args['column'])
else:
bfactors = zeros(len(points))
model = Model(1)
chain = Chain("A")
structure = Structure("ref")
num_count = 0
for i in range(0,shape(points)[0]):
num_count = num_count +1
res_id = (' ',num_count,' ')
residue = Residue(res_id,'ALA',' ')
cur_coord = tuple(points[i])
bfactor = bfactors[i]
atom = Atom('CA',cur_coord,bfactor,0,' ','CA',num_count,'C')
residue.add(atom)
chain.add(residue)
model.add(chain)
points[ind] = array((line.split())[0:3])
points[ind] = points[ind] * scale
ind = ind+1
return points
# --------------------------------------------------------------------
num_nodes = 1
# points = EnumerateSphereSpace(array([0, 0, 0]),length=4)
# start_point_1 = (1.236 , 0.000 , 3.804)
# start_point_2 = (1.236 , 3.804 , 5.040)
# start_point = start_point_2
# points = EnumerateSphereSpace(array([start_point]),length=4)
points = ReadXYZ(ptsfilename,scale)
# points = EnumerateSphereSpace()
model = Model(1)
chain=[]
chain.append (Chain("A"))
chain.append (Chain("B"))
chain.append (Chain("C"))
chain.append (Chain("D"))
chain.append (Chain("E"))
chain.append (Chain("F"))
chain.append (Chain("G"))
chain.append (Chain("H"))
chain.append (Chain("I"))
chain.append (Chain("J"))
chain.append (Chain("K"))
chain.append (Chain("L"))
chain.append (Chain("M"))
num_lines = sum(1 for line in open(filename))
points = zeros(shape=(num_lines,3))
ind = 0
for line in open(filename).readlines():
points[ind] = array((line.split()[0:3]))
points[ind] = points[ind] * scale
ind = ind+1
return points
#--------------------------------------------------------------------
ref_ptsfilename = "K562.pts"
refid = "ref"
structure = Structure(refid)
model_ref = Model(1)
chain_ref = Chain("A")
points_ref = ReadXYZ(ref_ptsfilename,scale)
num_count = 0
for i in range(0,shape(points_ref[IndexList])[0]):
num_count = num_count +1
res_id = (' ',num_count,' ')
residue = Residue(res_id,'ALA',' ')
cur_coord = tuple(points_ref[IndexList[i]])
atom = Atom('CA',cur_coord,0,0,' ',num_count,num_count,'C')
residue.add(atom)
chain_ref.add(residue)
model_ref.add(chain_ref)
structure.add(model_ref)
def run(self, struct: Model) -> List[Chain]:
return list(
filter(
lambda chain: sum(is_aa(residue) for residue in chain) >= 50,
struct.get_chains()))
class StructureBuilder:
"""Deals with constructing the Structure object.
The StructureBuilder class is used by the PDBParser classes to
translate a file to a Structure object.
"""
def __init__(self):
"""Initialize the class."""
self.line_counter = 0
self.header = {}
def _is_completely_disordered(self, residue):
"""Return 1 if all atoms in the residue have a non blank altloc (PRIVATE)."""
atom_list = residue.get_unpacked_list()
for atom in atom_list:
altloc = atom.get_altloc()
if altloc == " ":
return 0
return 1
# Public methods called by the Parser classes
def set_header(self, header):
"""Set header."""
self.header = header
def set_line_counter(self, line_counter):
"""Tracks line in the PDB file that is being parsed.
Arguments:
- line_counter - int
"""
self.line_counter = line_counter
def init_structure(self, structure_id):
"""Initialize a new Structure object with given id.
Arguments:
- id - string
"""
self.structure = Structure(structure_id)
def init_model(self, model_id, serial_num=None):
"""Create a new Model object with given id.
Arguments:
- id - int
- serial_num - int
"""
self.model = Model(model_id, serial_num)
self.structure.add(self.model)
def init_chain(self, chain_id):
"""Create a new Chain object with given id.
Arguments:
- chain_id - string
"""
if self.model.has_id(chain_id):
self.chain = self.model[chain_id]
warnings.warn(
"WARNING: Chain %s is discontinuous at line %i."
% (chain_id, self.line_counter),
PDBConstructionWarning,
)
else:
self.chain = Chain(chain_id)
self.model.add(self.chain)
def init_seg(self, segid):
"""Flag a change in segid.
Arguments:
- segid - string
"""
self.segid = segid
def init_residue(self, resname, field, resseq, icode):
"""Create a new Residue object.
Arguments:
- resname - string, e.g. "ASN"
- field - hetero flag, "W" for waters, "H" for
hetero residues, otherwise blank.
- resseq - int, sequence identifier
- icode - string, insertion code
"""
if field != " ":
if field == "H":
# The hetero field consists of H_ + the residue name (e.g. H_FUC)
field = "H_" + resname
res_id = (field, resseq, icode)
if field == " ":
if self.chain.has_id(res_id):
# There already is a residue with the id (field, resseq, icode).
# This only makes sense in the case of a point mutation.
warnings.warn(
"WARNING: Residue ('%s', %i, '%s') redefined at line %i."
% (field, resseq, icode, self.line_counter),
PDBConstructionWarning,
)
duplicate_residue = self.chain[res_id]
if duplicate_residue.is_disordered() == 2:
# The residue in the chain is a DisorderedResidue object.
# So just add the last Residue object.
if duplicate_residue.disordered_has_id(resname):
# The residue was already made
self.residue = duplicate_residue
duplicate_residue.disordered_select(resname)
else:
# Make a new residue and add it to the already
# present DisorderedResidue
new_residue = Residue(res_id, resname, self.segid)
duplicate_residue.disordered_add(new_residue)
self.residue = duplicate_residue
return
else:
if resname == duplicate_residue.resname:
warnings.warn(
"WARNING: Residue ('%s', %i, '%s','%s') already defined "
"with the same name at line %i."
% (field, resseq, icode, resname, self.line_counter),
PDBConstructionWarning,
)
self.residue = duplicate_residue
return
# Make a new DisorderedResidue object and put all
# the Residue objects with the id (field, resseq, icode) in it.
# These residues each should have non-blank altlocs for all their atoms.
# If not, the PDB file probably contains an error.
if not self._is_completely_disordered(duplicate_residue):
# if this exception is ignored, a residue will be missing
self.residue = None
raise PDBConstructionException(
"Blank altlocs in duplicate residue %s ('%s', %i, '%s')"
% (resname, field, resseq, icode)
)
self.chain.detach_child(res_id)
new_residue = Residue(res_id, resname, self.segid)
disordered_residue = DisorderedResidue(res_id)
self.chain.add(disordered_residue)
disordered_residue.disordered_add(duplicate_residue)
disordered_residue.disordered_add(new_residue)
self.residue = disordered_residue
return
self.residue = Residue(res_id, resname, self.segid)
self.chain.add(self.residue)
def init_atom(
self,
name,
coord,
b_factor,
occupancy,
altloc,
fullname,
serial_number=None,
element=None,
pqr_charge=None,
radius=None,
is_pqr=False,
):
"""Create a new Atom object.
Arguments:
- name - string, atom name, e.g. CA, spaces should be stripped
- coord - Numeric array (Float0, size 3), atomic coordinates
- b_factor - float, B factor
- occupancy - float
- altloc - string, alternative location specifier
- fullname - string, atom name including spaces, e.g. " CA "
- element - string, upper case, e.g. "HG" for mercury
- pqr_charge - float, atom charge (PQR format)
- radius - float, atom radius (PQR format)
- is_pqr - boolean, flag to specify if a .pqr file is being parsed
"""
residue = self.residue
# if residue is None, an exception was generated during
# the construction of the residue
if residue is None:
return
# First check if this atom is already present in the residue.
# If it is, it might be due to the fact that the two atoms have atom
# names that differ only in spaces (e.g. "CA.." and ".CA.",
# where the dots are spaces). If that is so, use all spaces
# in the atom name of the current atom.
if residue.has_id(name):
duplicate_atom = residue[name]
# atom name with spaces of duplicate atom
duplicate_fullname = duplicate_atom.get_fullname()
if duplicate_fullname != fullname:
# name of current atom now includes spaces
name = fullname
warnings.warn(
"Atom names %r and %r differ only in spaces at line %i."
% (duplicate_fullname, fullname, self.line_counter),
PDBConstructionWarning,
)
if not is_pqr:
self.atom = Atom(
name,
coord,
b_factor,
occupancy,
altloc,
fullname,
serial_number,
element,
)
elif is_pqr:
self.atom = Atom(
name,
coord,
None,
None,
altloc,
fullname,
serial_number,
element,
pqr_charge,
radius,
)
if altloc != " ":
# The atom is disordered
if residue.has_id(name):
# Residue already contains this atom
duplicate_atom = residue[name]
if duplicate_atom.is_disordered() == 2:
duplicate_atom.disordered_add(self.atom)
else:
# This is an error in the PDB file:
# a disordered atom is found with a blank altloc
# Detach the duplicate atom, and put it in a
# DisorderedAtom object together with the current
# atom.
residue.detach_child(name)
disordered_atom = DisorderedAtom(name)
residue.add(disordered_atom)
disordered_atom.disordered_add(self.atom)
disordered_atom.disordered_add(duplicate_atom)
residue.flag_disordered()
warnings.warn(
"WARNING: disordered atom found with blank altloc before "
"line %i.\n" % self.line_counter,
PDBConstructionWarning,
)
else:
# The residue does not contain this disordered atom
# so we create a new one.
disordered_atom = DisorderedAtom(name)
residue.add(disordered_atom)
# Add the real atom to the disordered atom, and the
# disordered atom to the residue
disordered_atom.disordered_add(self.atom)
residue.flag_disordered()
else:
# The atom is not disordered
residue.add(self.atom)
def set_anisou(self, anisou_array):
"""Set anisotropic B factor of current Atom."""
self.atom.set_anisou(anisou_array)
def set_siguij(self, siguij_array):
"""Set standard deviation of anisotropic B factor of current Atom."""
self.atom.set_siguij(siguij_array)
def set_sigatm(self, sigatm_array):
"""Set standard deviation of atom position of current Atom."""
self.atom.set_sigatm(sigatm_array)
def get_structure(self):
"""Return the structure."""
# first sort everything
# self.structure.sort()
# Add the header dict
self.structure.header = self.header
return self.structure
def set_symmetry(self, spacegroup, cell):
"""Set symmetry."""
pass
# nie jestem pewien dlaczego to nie działa. Pojawia się błąd:
# AttributeError: 'module' object has no attribute 'array'
from Bio import PDB
from Bio.PDB import PDBParser, NeighborSearch, Superimposer, PDBIO
from Bio.PDB.Atom import Atom
from Bio.PDB.Residue import Residue
from Bio.PDB.Chain import Chain
from Bio.PDB.Model import Model
from Bio.PDB.Structure import Structure
my_structure = Structure('Cytosine')
my_model = Model(0)
my_chain = Chain('A')
my_residue = Residue((' ', 1, ' '), 'C', ' ')
atoms = [{
'name': 'N1',
'coord': PDB.Atom.array([64.612, 45.818, 10.877], 'f'),
'bfactor': 42.59,
'occupancy': 1.0,
'altloc': ' ',
'fullname': 'N1',
'serial_number': 1
}, {
'name': 'C2',
'coord': PDB.Atom.array([65.472, 46.868, 10.634], 'f'),
'bfactor': 44.48,
'occupancy': 1.0,
'altloc': ' ',
'fullname': 'C2',
'serial_number': 2
