def find_bb_hbonds(residues):
cutoff_d_of_n_o = 3.5
vertices = []
atoms = []
for i_residue, residue in enumerate(residues):
residue.i = i_residue
for atom in residue.atoms():
atom.residue = residue
if residue.has_atom('O'):
atom = residue.atom('O')
atoms.append(atom)
vertices.append(atom.pos)
if residue.has_atom('N'):
atom = residue.atom('N')
atoms.append(atom)
vertices.append(atom.pos)
for i, j in spacehash.SpaceHash(vertices).close_pairs():
if abs(i - j) < 3:
continue
if atoms[i].element == 'O' and atoms[j].element == 'N':
o = atoms[i]
n = atoms[j]
elif atoms[i].element == 'N' and atoms[j].element == 'O':
n = atoms[i]
o = atoms[j]
else:
continue
if v3.distance(o.pos, n.pos) < cutoff_d_of_n_o:
o.residue.co_partners.append(n.residue.i)
n.residue.nh_partners.append(o.residue.i)
def get_width(atoms, center):
max_diff = 0
for atom in atoms:
diff = v3.distance(atom.pos, center)
if diff > max_diff:
max_diff = diff
return 2*max_diff
def find_bb_hbonds(residues):
cutoff_d_of_n_o = 3.5
vertices = []
atoms = []
for i_residue, residue in enumerate(residues):
residue.i = i_residue
for atom in residue.atoms():
atom.residue = residue
if residue.has_atom('O'):
atom = residue.atom('O')
atoms.append(atom)
vertices.append(atom.pos)
if residue.has_atom('N'):
atom = residue.atom('N')
atoms.append(atom)
vertices.append(atom.pos)
for i, j in spacehash.SpaceHash(vertices).close_pairs():
if abs(i - j) < 3:
continue
if atoms[i].element == 'O' and atoms[j].element == 'N':
o = atoms[i]
n = atoms[j]
elif atoms[i].element == 'N' and atoms[j].element == 'O':
n = atoms[i]
o = atoms[j]
else:
continue
if v3.distance(o.pos, n.pos) < cutoff_d_of_n_o:
o.residue.co_partners.append(n.residue.i)
n.residue.nh_partners.append(o.residue.i)
def get_width(atoms, center):
max_diff = 0
for atom in atoms:
diff = v3.distance(atom.pos, center)
if diff > max_diff:
max_diff = diff
return 2 * max_diff
def is_peptide_connected(res1, res2):
if res1.has_atom('CA') and \
res1.has_atom('C') and \
res2.has_atom('N') and \
res2.has_atom('CA'):
d = v3.distance(res1.atom('C').pos, res2.atom('N').pos)
if d < 2.0:
return True
return False
def is_peptide_connected(res1, res2):
if res1.has_atom('CA') and \
res1.has_atom('C') and \
res2.has_atom('N') and \
res2.has_atom('CA'):
d = v3.distance(res1.atom('C').pos, res2.atom('N').pos)
if d < 2.0:
return True
return False
def is_connected(i, j, soup, cutoff=3.5):
if i == j:
return False
min_dist = 1000.0
for atom_i in soup.residue(i).atoms():
for atom_j in soup.residue(j).atoms():
dist = v3.distance(atom_i.pos, atom_j.pos)
if dist < min_dist:
min_dist = dist
return min_dist < cutoff
def sum_rmsd(crds1, crds2):
"""
Returns the direct rmsd between two sets of vectors *without*
doing any optimal rotation. If calculated between optimal sets,
should give the proper RMSD.
"""
sum_squared = 0.0
for crd1, crd2 in zip(crds1, crds2):
sum_squared += v3.distance(crd1, crd2)**2
return math.sqrt(sum_squared/float(len(crds1)))
def is_connected(i, j, soup, cutoff=3.5):
if i == j:
return False
min_dist = 1000.0
for atom_i in soup.residue(i).atoms():
for atom_j in soup.residue(j).atoms():
dist = v3.distance(atom_i.pos, atom_j.pos)
if dist < min_dist:
min_dist = dist
return min_dist < cutoff
def sum_rmsd(crds1, crds2):
"""
Returns the direct rmsd between two sets of vectors *without*
doing any optimal rotation. If calculated between optimal sets,
should give the proper RMSD.
"""
sum_squared = 0.0
for crd1, crd2 in zip(crds1, crds2):
sum_squared += v3.distance(crd1, crd2)**2
return math.sqrt(sum_squared / float(len(crds1)))
def get_width(atoms, center=None):
"""
Returns twice the longest distance from the center.
"""
max_diff = 0
if center is None:
center = get_center(atoms)
for atom in atoms:
diff = v3.distance(atom.pos, center)
if diff > max_diff:
max_diff = diff
return 2*max_diff
def is_sidechain_connected(i, j, soup, cutoff=3.5):
if abs(i-j) <= 2:
return False
min_dist = 1000.0
sidechain_atoms_i = [a for a in soup.residue(i).atoms()
if a.type not in backbone]
for atom_i in sidechain_atoms_i:
for atom_j in soup.residue(j).atoms():
dist = v3.distance(atom_i.pos, atom_j.pos)
if dist < min_dist:
min_dist = dist
return min_dist < cutoff
def is_sidechain_connected(i, j, soup, cutoff=3.5):
if abs(i - j) <= 2:
return False
min_dist = 1000.0
sidechain_atoms_i = [
a for a in soup.residue(i).atoms() if a.type not in backbone
]
for atom_i in sidechain_atoms_i:
for atom_j in soup.residue(j).atoms():
dist = v3.distance(atom_i.pos, atom_j.pos)
if dist < min_dist:
min_dist = dist
return min_dist < cutoff
def find_neighbor_indices(atoms, probe, k):
"""
Returns list of indices of atoms within probe distance to atom k.
"""
neighbor_indices = []
atom_k = atoms[k]
radius = atom_k.radius + probe + probe
indices = range(k)
indices.extend(range(k + 1, len(atoms)))
for i in indices:
atom_i = atoms[i]
dist = v3.distance(atom_k.pos, atom_i.pos)
if dist < radius + atom_i.radius:
neighbor_indices.append(i)
return neighbor_indices
def find_neighbor_indices(atoms, probe, k):
"""
Returns list of indices of atoms within probe distance to atom k.
"""
neighbor_indices = []
atom_k = atoms[k]
radius = atom_k.radius + probe + probe
indices = range(k)
indices.extend(range(k+1, len(atoms)))
for i in indices:
atom_i = atoms[i]
dist = v3.distance(atom_k.pos, atom_i.pos)
if dist < radius + atom_i.radius:
neighbor_indices.append(i)
return neighbor_indices
def calculate_results(self):
ref_crds = []
crds = []
for ref_residue, residue in zip(self.ref_soup.residues(), self.soup.residues()):
if residue.type not in data.solvent_res_types:
if residue.has_atom('CA'):
ref_crds.append(ref_residue.atom('CA').pos.copy())
crds.append(residue.atom('CA').pos.copy())
center = v3.get_center(crds)
crds = [c - center for c in crds]
ref_center = v3.get_center(ref_crds)
ref_crds = [c - ref_center for c in ref_crds]
rmsd_val, transform_ref_to_this = rmsd.calc_rmsd_rot(ref_crds, crds)
ref_crds = [v3.transform(transform_ref_to_this, c) for c in ref_crds]
return [v3.distance(crd, ref_crd) for crd, ref_crd in zip(crds, ref_crds)]
def calculate_results(self):
ref_crds = []
crds = []
for ref_residue, residue in zip(self.ref_soup.residues(),
self.soup.residues()):
if residue.type not in data.solvent_res_types:
if residue.has_atom('CA'):
ref_crds.append(ref_residue.atom('CA').pos.copy())
crds.append(residue.atom('CA').pos.copy())
center = v3.get_center(crds)
crds = [c - center for c in crds]
ref_center = v3.get_center(ref_crds)
ref_crds = [c - ref_center for c in ref_crds]
rmsd_val, transform_ref_to_this = rmsd.calc_rmsd_rot(ref_crds, crds)
ref_crds = [v3.transform(transform_ref_to_this, c) for c in ref_crds]
return [
v3.distance(crd, ref_crd) for crd, ref_crd in zip(crds, ref_crds)
]
def calculate_asa(atoms, probe, n_sphere_point=960):
"""
Returns the accessible-surface areas of the atoms, by rolling a
ball with probe radius over the atoms with their radius
defined.
"""
sphere_points = generate_sphere_points(n_sphere_point)
const = 4.0 * math.pi / len(sphere_points)
areas = []
for i, atom_i in enumerate(atoms):
neighbor_indices = find_neighbor_indices(atoms, probe, i)
n_neighbor = len(neighbor_indices)
j_closest_neighbor = 0
radius = probe + atom_i.radius
n_accessible_point = 0
for point in sphere_points:
is_accessible = True
test_point = v3.scale(point, radius) + atom_i.pos
cycled_indices = range(j_closest_neighbor, n_neighbor)
cycled_indices.extend(range(j_closest_neighbor))
for j in cycled_indices:
atom_j = atoms[neighbor_indices[j]]
r = atom_j.radius + probe
diff = v3.distance(atom_j.pos, test_point)
if diff * diff < r * r:
j_closest_neighbor = j
is_accessible = False
break
if is_accessible:
n_accessible_point += 1
area = const * n_accessible_point * radius * radius
areas.append(area)
return areas
def calculate_asa(atoms, probe, n_sphere_point=960):
"""
Returns the accessible-surface areas of the atoms, by rolling a
ball with probe radius over the atoms with their radius
defined.
"""
sphere_points = generate_sphere_points(n_sphere_point)
const = 4.0 * math.pi / len(sphere_points)
areas = []
for i, atom_i in enumerate(atoms):
neighbor_indices = find_neighbor_indices(atoms, probe, i)
n_neighbor = len(neighbor_indices)
j_closest_neighbor = 0
radius = probe + atom_i.radius
n_accessible_point = 0
for point in sphere_points:
is_accessible = True
test_point = v3.scale(point, radius) + atom_i.pos
cycled_indices = range(j_closest_neighbor, n_neighbor)
cycled_indices.extend(range(j_closest_neighbor))
for j in cycled_indices:
atom_j = atoms[neighbor_indices[j]]
r = atom_j.radius + probe
diff = v3.distance(atom_j.pos, test_point)
if diff*diff < r*r:
j_closest_neighbor = j
is_accessible = False
break
if is_accessible:
n_accessible_point += 1
area = const*n_accessible_point*radius*radius
areas.append(area)
return areas
def make_disulfide_script(pdb):
"""
Returns the psfgen script for disulfide bonds.
This function opens in_pdb in a soup object, and searches for
CYS residues where the SG-SG distance < 3 angs. These residues
are then renamed to CYX and written to out_pdb. The disulfide bonds
are then returned in a .tleap script fragment.
"""
soup = pdbatoms.Soup(pdb)
n = len(soup.residues())
# First generate the residue names recognized by psfgen
res_names = []
chain_id = None
i_res = None
for i in range(n):
res = soup.residue(i)
if res.chain_id != chain_id:
chain_id = res.chain_id
i_res = 1
res_names.append("%s:%s" % (chain_id, i_res))
i_res += 1
# Then search through for all CYS-CYS pairs and identify disulfide bonds
script = ""
for i in range(n):
for j in range(i+1, n):
if soup.residue(i).type in 'CYS' and soup.residue(j).type in 'CYS':
sg1 = soup.residue(i).atom('SG')
sg2 = soup.residue(j).atom('SG')
if v3.distance(sg1.pos, sg2.pos) < 3.0:
script += "patch DISU %s %s\n" % (res_names[i], res_names[j])
if script:
script = "# disulfide bonds\n" + script + "\n"
return script
def disulfide_script_and_rename_cysteines(in_pdb, out_pdb):
"""
Returns the tleap script for disulfide bonds in the in_pdb file.
This function opens in_pdb in a Soup, and searches for
CYS residues where the SG-SG distance < 3 angs. These residues
are then renamed to CYX and written to out_pdb. The disulfide bonds
are then returned in a .tleap script fragment.
"""
soup = pdbatoms.Soup(in_pdb)
script = " # disulfide bonds\n"
n = len(soup.residues())
for i in range(n):
for j in range(i + 1, n):
if soup.residue(i).type in 'CYS' and soup.residue(j).type in 'CYS':
p1 = soup.residue(i).atom('SG').pos
p2 = soup.residue(j).atom('SG').pos
if v3.distance(p1, p2) < 3.0:
soup.residue(i).set_type('CYX')
soup.residue(j).set_type('CYX')
script += "bond pdb.%d.SG pdb.%d.SG\n" % (i + 1, j + 1)
soup.write_pdb(out_pdb)
util.check_output(out_pdb)
return script
def disulfide_script_and_rename_cysteines(in_pdb, out_pdb):
"""
Returns the tleap script for disulfide bonds in the in_pdb file.
This function opens in_pdb in a Soup, and searches for
CYS residues where the SG-SG distance < 3 angs. These residues
are then renamed to CYX and written to out_pdb. The disulfide bonds
are then returned in a .tleap script fragment.
"""
soup = pdbatoms.Soup(in_pdb)
script = " # disulfide bonds\n"
n = len(soup.residues())
for i in range(n):
for j in range(i+1, n):
if soup.residue(i).type in 'CYS' and soup.residue(j).type in 'CYS':
p1 = soup.residue(i).atom('SG').pos
p2 = soup.residue(j).atom('SG').pos
if v3.distance(p1, p2) < 3.0:
soup.residue(i).set_type('CYX')
soup.residue(j).set_type('CYX')
script += "bond pdb.%d.SG pdb.%d.SG\n" % (i+1, j+1)
soup.write_pdb(out_pdb)
util.check_output(out_pdb)
return script
