def apply_rotations_to_chains(final_coordinates_atomistic,
atomistic_protein_centered, at_com_group,
cg_com_group, cg_com):
final_user_supplied_coord = {}
for chain in range(len(final_coordinates_atomistic)):
rotations = []
if chain in atomistic_protein_centered:
rotate_chain = at_mod.kabsch_rotate(
at_com_group[g_var.group_chains[chain]] -
cg_com[g_var.group_chains[chain]][0],
cg_com_group[g_var.group_chains[chain]] -
cg_com[g_var.group_chains[chain]][0])
for part_val, part in enumerate(atomistic_protein_centered[chain]):
if chain in g_var.group_chains:
rotations.append(rotate_chain)
else:
rotations = at_com_group[chain]
final_user_supplied_coord[chain] = hybridise_protein_inputs(
final_coordinates_atomistic[chain],
atomistic_protein_centered[chain], cg_com[chain], rotations,
chain)
else:
final_user_supplied_coord[chain] = hybridise_protein_inputs(
final_coordinates_atomistic[chain], [], [], [], chain)
return final_user_supplied_coord
def RMSD_align(coord_set_1, coord_set_2):
center = np.mean(coord_set_2, axis=0)
xyz_rot_apply=at_mod.kabsch_rotate(coord_set_1-center, coord_set_2-center)
ali= []
for at_val, atom in enumerate(coord_set_1):
ali.append( at_mod.rotate_atom(atom, center, xyz_rot_apply) )
return np.array(ali)
def return_all_rotations_final(at_com_group, cg_com_group, cg_com):
if len(at_com_group['all']) == len(cg_com_group['all']):
return at_mod.kabsch_rotate(at_com_group['all'] - cg_com[0][0],
cg_com_group['all'] - cg_com[0][0])
else:
sys.exit('The atomistic input does not match the CG. \n\
number of CG residues ' + str(len(cg_com_group['all'])) +
'\nnumber of AT residues ' + str(len(at_com_group['all'])))
def return_indivdual_rotations(chain, part_val, at_centers, cg_com, at_com_group, cg_com_group, sls, sle):
if chain not in at_com_group:
at_com_group[chain]=[]
if len(at_centers) == len(np.array(g_var.backbone_coords[chain])[sls:sle,:3]):
at_com_group[chain].append(at_mod.kabsch_rotate(np.array(at_centers)-cg_com[chain][part_val],
np.array(g_var.backbone_coords[chain])[sls:sle,:3]-cg_com[chain][part_val]))
else:
sys.exit('In chain '+str(chain)+' the atomistic input does not match the CG. \n\
number of CG residues '+str(len(g_var.backbone_coords[chain]))+'\nnumber of AT residues '+str(len(at_centers)))
return at_com_group
def RMSD_measure(structure_atoms):
RMSD_dict = {}
for chain in range(g_var.system['PROTEIN']):
at_centers = []
#### runs through every residue and atom
for residue in structure_atoms[chain]:
#### gets center of mass of each residue (note only backbone heavy atoms have a mass)
at_centers_iter = []
for atom in structure_atoms[chain][residue]:
at_centers_iter.append(
np.append(
structure_atoms[chain][residue][atom]['coord'],
structure_atoms[chain][residue][atom]['frag_mass']))
try:
at_centers.append(
np.average(np.array(at_centers_iter)[:, :3],
axis=0,
weights=np.array(at_centers_iter)[:, 3]))
except BaseException:
print('The fragment probably has no mass\n')
for atom in structure_atoms[chain][residue]:
print(structure_atoms[chain][residue][atom])
sys.exit()
#### checks that the number of residues in the chain are the same between CG and AT
if len(at_centers) != len(g_var.backbone_coords[chain]):
sys.exit('In chain ' + str(chain) +
' the atomistic input does not match the CG. \n\
number of CG residues ' + str(len(g_var.backbone_coords[chain])) +
'\nnumber of AT residues ' + str(len(at_centers)))
cg_center = np.mean(np.array(g_var.backbone_coords[chain])[:, :3],
axis=0)
at_align = np.array(at_centers) - (
np.mean(np.array(at_centers), axis=0) - cg_center)
xyz_rot_apply = at_mod.kabsch_rotate(
np.array(at_align) - cg_center,
np.array(np.array(g_var.backbone_coords[chain])[:, :3]) -
cg_center)
for at_val, atom in enumerate(at_align):
at_align[at_val] = at_mod.rotate_atom(atom, cg_center,
xyz_rot_apply)
#### finds distance between backbone COM and cg backbone beads
dist = np.sqrt((np.array(at_align) -
np.array(g_var.backbone_coords[chain])[:, :3])**2)
RMSD_val = np.sqrt(np.mean(dist**2)) #### RMSD calculation
RMSD_dict[chain] = np.round(RMSD_val,
3) #### stores RMSD in dictionary
return RMSD_dict
def atomistic_non_protein_non_solvent(cg_residue_type, cg_residues):
atomistic_fragments = {} #### residue dictionary
#### run through every residue in a particular residue type
residue_type = {}
residue_type_mass = {}
if not os.path.exists(g_var.working_dir + cg_residue_type + '/' +
cg_residue_type + '_all.pdb'):
for cg_resid, cg_residue in enumerate(cg_residues):
atomistic_fragments[cg_resid] = {}
frag_location = gen.fragment_location(
cg_residue_type) ### get fragment location from database
residue_type[cg_residue_type], residue_type_mass[
cg_residue_type] = at_mod.get_atomistic(frag_location)
for group in residue_type[cg_residue_type]:
center, at_frag_centers, cg_frag_centers, group_fit = at_mod.rigid_fit(
residue_type[cg_residue_type][group],
residue_type_mass[cg_residue_type], cg_residue,
cg_residues[cg_residue])
at_connect, cg_connect = at_mod.connectivity(
cg_residues[cg_residue], at_frag_centers, cg_frag_centers,
group_fit, group)
if len(at_connect) == len(cg_connect) and len(cg_connect) > 0:
try:
xyz_rot_apply = at_mod.kabsch_rotate(
np.array(at_connect) - center,
np.array(cg_connect) - center)
except BaseException:
sys.exit('There is a issue with residue: ' +
cg_residue_type + ' in group: ' + str(group))
else:
print('atom connections: ' + str(len(at_connect)) +
' does not match CG connections: ' +
str(len(cg_connect)))
sys.exit('residue number: ' + str(cg_resid) +
', residue type: ' + str(cg_residue_type) +
', group: ' + group)
for bead in group_fit:
for atom in group_fit[bead]:
group_fit[bead][atom]['coord'] = at_mod.rotate_atom(
group_fit[bead][atom]['coord'], center,
xyz_rot_apply)
atom_new = group_fit[bead][atom].copy()
atomistic_fragments[cg_resid][atom] = atom_new
return atomistic_fragments, 0
else:
return atomistic_fragments, len(cg_residues)
def build_multi_residue_atomistic_system(cg_residues, sys_type):
#### initisation of counters
chain_count = 0
coord_atomistic = {}
g_var.seq_cg = {sys_type: {}}
g_var.ter_res = {sys_type: {}}
gen.mkdir_directory(g_var.working_dir +
sys_type) ### make and change to protein directory
#### for each residue in protein
residue_type = {}
residue_type_mass = {}
new_chain = True
for cg_residue_id, residue_number in enumerate(cg_residues[sys_type]):
if np.round((cg_residue_id / len(cg_residues[sys_type])) * 100,
2).is_integer():
print('Converting de_novo ' + sys_type + ': ',
np.round((cg_residue_id / len(cg_residues[sys_type])) * 100,
2),
'%',
end='\r')
resname = cg_residues[sys_type][residue_number][next(
iter(cg_residues[sys_type][residue_number]))]['residue_name']
if new_chain:
if chain_count not in coord_atomistic:
if sys_type == 'PROTEIN':
g_var.backbone_coords[chain_count] = []
coord_atomistic[chain_count] = {}
g_var.seq_cg[sys_type][chain_count] = []
g_var.ter_res[sys_type][chain_count] = [resname, False]
new_chain = False
coord_atomistic[chain_count][residue_number] = {}
frag_location = gen.fragment_location(
resname) ### get fragment location from database
residue_type[resname], residue_type_mass[
resname] = at_mod.get_atomistic(frag_location)
g_var.seq_cg[sys_type] = add_to_sequence(g_var.seq_cg[sys_type],
resname, chain_count)
new_chain = False
for group in residue_type[resname]:
for key in list(residue_type[resname][group].keys()):
if key not in cg_residues[sys_type][residue_number]:
del residue_type[resname][group][key]
if len(residue_type[resname][group]) > 0:
center, at_frag_centers, cg_frag_centers, group_fit = at_mod.rigid_fit(
residue_type[resname][group], residue_type_mass[resname],
residue_number, cg_residues[sys_type][residue_number])
at_connect, cg_connect = at_mod.connectivity(
cg_residues[sys_type][residue_number], at_frag_centers,
cg_frag_centers, group_fit, group)
for group_bead in group_fit:
if group_bead in g_var.res_top[resname]['CONNECT']:
at_connect, cg_connect, new_chain = at_mod.BB_connectivity(
at_connect, cg_connect, cg_residues[sys_type],
group_fit[group_bead], residue_number, group_bead)
if sys_type == 'PROTEIN':
g_var.backbone_coords[chain_count].append(
np.append(
cg_residues[sys_type][residue_number]
[group_bead]['coord'], 1))
if len(at_connect) == len(cg_connect) and len(at_connect) != 0:
xyz_rot_apply = at_mod.kabsch_rotate(
np.array(at_connect) - center,
np.array(cg_connect) - center)
elif len(at_connect) == 0:
xyz_rot_apply = False
print('Cannot find any connectivity for residue number: ' +
str(residue_number) + ', residue type: ' +
str(resname) + ', group: ' + str(group))
else:
print('atom connections: ' + str(len(at_connect)) +
' does not equal CG connections: ' +
str(len(cg_connect)))
sys.exit('residue number: ' + str(residue_number) +
', residue type: ' + str(resname) + ', group: ' +
group)
for bead in group_fit:
for atom in group_fit[bead]:
group_fit[bead][atom]['coord'] = at_mod.rotate_atom(
group_fit[bead][atom]['coord'], center,
xyz_rot_apply)
atom_new = group_fit[bead][atom].copy()
coord_atomistic[chain_count][residue_number][
atom] = atom_new
if new_chain:
g_var.ter_res[sys_type][chain_count][1] = resname
chain_count += 1
print('Completed initial conversion of ' + sys_type + '\n')
g_var.system[sys_type] = chain_count
if sys_type == 'PROTEIN':
for chain in range(chain_count):
g_var.skip_disul[chain] = False
return coord_atomistic