Bio Structural Biology Structure Modification

from Bio.PDB import PDBParser, PDBIO
import numpy as np

parser = PDBParser(QUIET=True)
structure = parser.get_structure('protein', 'protein.pdb')

# Entity.transform applies coords @ rot + tran (row convention) to every atom in place.
identity = np.identity(3)
translation = np.array([10.0, 0.0, 0.0])
structure.transform(identity, translation)

io = PDBIO()
io.set_structure(structure)
io.save('translated.pdb')

from Bio.PDB import PDBParser
from Bio.PDB.vectors import rotaxis, Vector
import numpy as np

parser = PDBParser(QUIET=True)
structure = parser.get_structure('protein', 'protein.pdb')

# rotaxis returns a row-convention matrix intended for Entity/atom.transform.
rot = rotaxis(np.radians(90), Vector(0, 0, 1))

# Rotate about the center of mass: pick tran so the center is the fixed point of coords @ rot + tran.
center = np.array([a.coord for a in structure.get_atoms()]).mean(axis=0)
tran = center - center @ rot
structure.transform(rot, tran)

from Bio.PDB import PDBParser
import numpy as np

parser = PDBParser(QUIET=True)
structure = parser.get_structure('protein', 'protein.pdb')

# REMARK 350 / _pdbx_struct_oper_list operators are column-convention: newcoord = R @ coord + t.
R = np.array([[0.0, -1.0, 0.0], [1.0, 0.0, 0.0], [0.0, 0.0, 1.0]])
t = np.array([25.0, 0.0, 0.0])

# Entity.transform expects the row convention, so transpose the column-convention R first.
structure.transform(R.T, t)

from Bio.PDB import PDBParser
import numpy as np

parser = PDBParser(QUIET=True)
structure = parser.get_structure('protein', 'protein.pdb')

center = np.array([a.coord for a in structure.get_atoms()]).mean(axis=0)
structure.transform(np.identity(3), -center)

from Bio.PDB import PDBParser, PDBIO

parser = PDBParser(QUIET=True)
structure = parser.get_structure('protein', 'protein.pdb')
model = structure[0]

# Detach hydrogens; collect ids first so the child dict is not mutated mid-iteration.
for residue in model.get_residues():
    for atom_id in [a.id for a in residue if a.element == 'H']:
        residue.detach_child(atom_id)

# Detach whole chains by id.
if model.has_id('B'):
    model.detach_child('B')

io = PDBIO()
io.set_structure(structure)
io.save('cleaned.pdb')

from Bio.PDB import PDBParser, PDBIO

parser = PDBParser(QUIET=True)
structure = parser.get_structure('protein', 'protein.pdb')

# 'W' HETFLAG isolates water; a blanket r.id[0] != ' ' would also delete Zn/Mg/heme/FAD and MSE.
for chain in structure[0]:
    for res_id in [r.id for r in chain if r.id[0] == 'W']:
        chain.detach_child(res_id)

io = PDBIO()
io.set_structure(structure)
io.save('no_water.pdb')

from Bio.PDB import PDBParser, PDBIO, Select

parser = PDBParser(QUIET=True)
structure = parser.get_structure('protein', 'protein.pdb')

# Select writes a filtered copy without mutating the parsed tree.
class CoreChain(Select):
    def accept_chain(self, chain):
        return chain.id == 'A'
    def accept_residue(self, residue):
        return residue.id[0] == ' ' and 50 <= residue.id[1] <= 100

io = PDBIO()
io.set_structure(structure)
io.save('coreA_50_100.pdb', CoreChain())

from Bio.PDB import PDBParser, PDBIO

parser = PDBParser(QUIET=True)
structure = parser.get_structure('protein', 'protein.pdb')

# Snapshot originals: this column is a real temperature factor (or AlphaFold pLDDT) until overwritten.
original_bfactors = {atom.get_full_id(): atom.bfactor for atom in structure.get_atoms()}

conservation = {100: 9.0, 101: 5.0, 102: 3.0}
for residue in structure.get_residues():
    score = conservation.get(residue.id[1])
    if score is None:
        continue
    for atom in residue:
        atom.bfactor = score  # set on all atoms so per-atom coloring is not patchy

io = PDBIO()
io.set_structure(structure)
io.save('colored.pdb')  # do not feed this file back to refinement/validation

from Bio.PDB import PDBParser, PDBIO

parser = PDBParser(QUIET=True)
structure = parser.get_structure('protein', 'protein.pdb')

# Occupancy must stay consistent with altlocs: complementary altlocs should sum to <= 1.
for atom in structure[0]['A'].get_atoms():
    atom.occupancy = 1.0

io = PDBIO()
io.set_structure(structure)
io.save('occupancy_set.pdb')

from Bio.PDB import PDBParser, PDBIO

parser = PDBParser(QUIET=True)
structure = parser.get_structure('protein', 'protein.pdb')
chain = structure[0]['A']

# Preserve the (hetflag, ..., icode) tuple; only the resseq middle field changes.
# Assign into a temporary range first to avoid colliding with existing ids mid-loop.
for offset, residue in enumerate(list(chain)):
    hetflag, _, icode = residue.id
    residue.id = (hetflag, offset + 10000, icode)
for new_seq, residue in enumerate(list(chain), start=1):
    hetflag, _, icode = residue.id
    residue.id = (hetflag, new_seq, icode)

io = PDBIO()
io.set_structure(structure)
io.save('renumbered.pdb')