def fix_threaded_lipids(lipid_atoms, merge, merge_coords):
if not os.path.exists(g_var.merged_directory +
'merged_cg2at_threaded.pdb'):
tree = cKDTree(merge_coords)
for threaded in lipid_atoms:
resname = get_np_resname(threaded[0])
atoms = tree.query_ball_point(threaded[2], r=3)
for at in atoms:
if merge[at]['residue_id'] != merge[threaded[0]]['residue_id']:
P_count = fetch_start_of_residue(at, merge)
break
NP_count = fetch_start_of_residue_np(threaded[0], resname)
bb = []
if 'P_count' not in locals():
sys.exit('There is an issue with the bond length detection')
for at in merge[P_count:]:
if at['residue_id'] != merge[P_count]['residue_id']:
break
if at['atom_name'] in g_var.res_top[
at['residue_name']]['ATOMS']:
bb.append([at['x'], at['y'], at['z']])
bb = np.mean(np.array(bb), axis=0)
BB_M3 = (threaded[2] - bb) / np.linalg.norm((threaded[2] - bb))
for heavy_atom in threaded[:2]:
merge_coords[heavy_atom] += BB_M3 * 3
merge[heavy_atom]['x'], merge[heavy_atom]['y'], merge[
heavy_atom]['z'] = merge_coords[heavy_atom]
for hydrogen in g_var.hydrogen[resname][heavy_atom - NP_count +
1]:
merge_coords[NP_count + hydrogen - 1] += BB_M3 * 3
merge[NP_count + hydrogen -
1]['x'], merge[NP_count + hydrogen - 1]['y'], merge[
NP_count + hydrogen -
1]['z'] = merge_coords[NP_count + hydrogen - 1]
coords, index_conversion = index_conversion_generate(
merge, merge_coords)
write_pdb(merge, coords, index_conversion,
g_var.merged_directory + 'merged_cg2at_threaded.pdb')
if not os.path.exists(g_var.merged_directory +
'MIN/merged_cg2at_threaded_minimised.pdb'):
gro.minimise_merged_pdbs('_threaded')
gen.file_copy_and_check(
g_var.merged_directory + 'MIN/merged_cg2at_threaded_minimised.pdb',
g_var.merged_directory + 'checked_ringed_lipid_de_novo.pdb')
'high_posre.itp'
]):
if file_name.endswith('.itp') or file_name.endswith('final.top'):
gen.file_copy_and_check(g_var.merged_directory + file_name,
g_var.final_dir + file_name)
#### merges provided atomistic protein and residues types into a single pdb file into merged directory
if not os.path.exists(g_var.merged_directory + 'merged_cg2at_de_novo.pdb'):
at_mod.merge_system_pdbs(
'_de_novo') ## merge all minimised residues into a complete system
## minimise merged system
if not os.path.exists(g_var.merged_directory +
'MIN/merged_cg2at_de_novo_minimised.pdb'):
gro.make_min('merged_cg2at')
gro.minimise_merged_pdbs('_de_novo') ## minimise system pdb
g_var.tc['m_t'] = time.time()
## checks for threaded lipids, e.g. abnormal bonds lengths (not had a issue for a long time might delete)
if not os.path.exists(g_var.merged_directory +
'checked_ringed_lipid_de_novo.pdb'):
at_mod.check_ringed_lipids(g_var.merged_directory +
'MIN/merged_cg2at_de_novo_minimised.pdb'
) ## check for abnormal bond lengths
## runs short NVT on de_novo system with disres on if available
if g_var.args.o not in ['none', 'align']:
gro.run_nvt(g_var.merged_directory +
'checked_ringed_lipid_de_novo') ## run npt on system
else:
print(
'Completed initial minimisation, please find final de_novo system: \n'
+ g_var.final_dir + 'final_cg2at_de_novo.pdb')
#### creates merged folder
print('\nMerging all residue types to single file. (Or a possibly tea)\n')
if len(system) > 0:
gen.mkdir_directory(g_var.working_dir + 'MERGED')
#### make final topology in merged directory
gro.write_merged_topol(system, '_novo')
#### make minimisation directory
gro.make_min('merged_cg2at')
#### merges provided atomistic protein and residues types into a single pdb file into merged directory
if user_at_input and 'PROTEIN' in system:
at_mod.merge_system_pdbs(system, '_no_steered', cg_residues, box_vec)
at_mod.merge_system_pdbs(system, '_at_rep_user_supplied', cg_residues,
box_vec)
gro.minimise_merged_pdbs(system, '_at_rep_user_supplied')
if len(system) > 1:
gro.alchembed(system['PROTEIN'])
else:
copyfile(
g_var.working_dir +
'MERGED/min/merged_cg2at_at_rep_user_supplied_minimised.pdb',
g_var.final_dir + 'final_cg2at_at_rep_user_supplied.pdb')
copyfile(g_var.working_dir + 'MERGED/merged_cg2at_no_steered.pdb',
g_var.final_dir + 'final_cg2at_no_steered.pdb')
#### merges de novo protein and residues types into a single pdb file into merged directory
at_mod.merge_system_pdbs(system, '_novo', cg_residues, box_vec)
gro.minimise_merged_pdbs(system, '_novo')
copyfile('merged_cg2at_novo_minimised.pdb',
g_var.final_dir + 'final_cg2at_de_novo.pdb')
merge_time = np.array(gmtime()[3:6])
def CG2AT_run(user_at_input):
gen.flags_used()
time_counter = {}
time_counter['i_t'] = time.time()
print(
'\nThis script is now hopefully doing the following (Good luck):\n\nReading in your CG representation\n'
)
#### reads in CG file and separates into residue types
cg_residues, box_vec_initial = read_in.read_initial_cg_pdb()
#### box size update
if g_var.box != None:
box_vec, box_shift = gen.new_box_vec(box_vec_initial, g_var.box)
else:
box_vec = box_vec_initial
box_shift = np.array([0, 0, 0])
#### simple pbc fix and residue truncation if required
cg_residues = read_in.fix_pbc(cg_residues, box_vec_initial, box_vec,
box_shift)
#### checks if fragment database is correct
at_mod.sanity_check(cg_residues)
time_counter['r_i_t'] = time.time()
system = {}
### convert protein to atomistic representation
if 'PROTEIN' in cg_residues:
p_system, backbone_coords, final_coordinates_atomistic, sequence = at_mod_p.build_protein_atomistic_system(
cg_residues['PROTEIN'], box_vec)
system['PROTEIN'] = p_system['PROTEIN']
time_counter['p_d_n_t'] = time.time()
#### reads in user supplied atomistic structure
if user_at_input and 'PROTEIN' in system:
atomistic_protein_input = at_mod_p.read_in_atomistic(
g_var.input_directory + 'AT_input.pdb', system['PROTEIN'],
sequence, True) ## reads in user structure
atomistic_protein_centered, cg_com = at_mod_p.center_atomistic(
atomistic_protein_input,
backbone_coords) ## centers each monomer by center of mass
at_mod_p.rotate_protein_monomers(
atomistic_protein_centered, final_coordinates_atomistic,
backbone_coords, cg_com, box_vec) ## rigid fits each monomer
#### minimise each protein chain
gro.minimise_protein(system['PROTEIN'], p_system, user_at_input)
#### read in minimised de novo protein chains and merges chains
merge_de_novo = at_mod_p.read_in_protein_pdbs(
system['PROTEIN'], g_var.working_dir + 'PROTEIN/min/PROTEIN_novo',
'.pdb')
at_mod_p.write_merged_pdb(merge_de_novo, '_novo', box_vec)
#### runs steered MD on user supplied protein chains
if user_at_input and 'PROTEIN' in system:
print('\tRunning steered MD on input atomistic structure\n')
#### runs steered MD on atomistic structure on CA and CB atoms
for chain in range(system['PROTEIN']):
gro.steered_md_atomistic_to_cg_coord(chain)
#### read in minimised user supplied protein chains and merges chains
merge_at_user = at_mod_p.read_in_protein_pdbs(
system['PROTEIN'], g_var.working_dir +
'PROTEIN/steered_md/PROTEIN_at_rep_user_supplied', '.pdb')
at_mod_p.write_merged_pdb(merge_at_user, '_at_rep_user_supplied',
box_vec)
merge_at_user_no_steer = at_mod_p.read_in_protein_pdbs(
system['PROTEIN'],
g_var.working_dir + 'PROTEIN/PROTEIN_at_rep_user_supplied',
'_gmx.pdb')
at_mod_p.write_merged_pdb(merge_at_user_no_steer, '_no_steered',
box_vec)
time_counter['f_p_t'] = time.time()
#### converts non protein residues into atomistic (runs on all cores)
if len([
key
for value, key in enumerate(cg_residues) if key not in ['PROTEIN']
]) > 0:
np_system = {}
pool = mp.Pool(mp.cpu_count())
pool_process = pool.starmap_async(
at_mod_np.build_atomistic_system,
[(cg_residues, residue_type, box_vec) for residue_type in [
key for value, key in enumerate(cg_residues)
if key not in ['PROTEIN']
]]).get() ## minimisation grompp parallised
pool.close()
for residue_type in pool_process:
np_system.update(residue_type)
#### minimises each residue separately
print('\nThis may take some time....(probably time for a coffee)\n')
for residue_type in [
key for value, key in enumerate(cg_residues)
if key not in ['PROTEIN', 'ION']
]:
print('Minimising individual residues: ' + residue_type)
gro.non_protein_minimise(np_system[residue_type], residue_type)
at_mod_np.merge_minimised(residue_type, np_system, box_vec)
print('Minimising merged: ' + residue_type)
gro.minimise_merged(residue_type, np_system)
system.update(np_system)
time_counter['b_n_p_t'] = time.time()
time_counter['n_p_t'] = time.time()
#### creates merged folder
print('\nMerging all residue types to single file. (Or possibly tea)\n')
if len(system) > 0:
#### make final topology in merged directory
gro.write_merged_topol(system, '_novo')
#### make minimisation directory
gro.make_min('merged_cg2at')
#### merges provided atomistic protein and residues types into a single pdb file into merged directory
if user_at_input and 'PROTEIN' in system:
at_mod.merge_system_pdbs(system, '_no_steered', cg_residues,
box_vec)
at_mod.merge_system_pdbs(system, '_at_rep_user_supplied',
cg_residues, box_vec)
gro.minimise_merged_pdbs(system, '_at_rep_user_supplied')
if len(system) > 1 and g_var.alchembed:
gro.alchembed(system['PROTEIN'])
else:
gen.file_copy_and_check(
g_var.working_dir +
'MERGED/min/merged_cg2at_at_rep_user_supplied_minimised.pdb',
g_var.final_dir + 'final_cg2at_at_rep_user_supplied.pdb')
gen.file_copy_and_check(
g_var.working_dir + 'MERGED/merged_cg2at_no_steered.pdb',
g_var.final_dir + 'final_cg2at_no_steered.pdb')
#### merges de novo protein and residues types into a single pdb file into merged directory
at_mod.merge_system_pdbs(system, '_novo', cg_residues, box_vec)
gro.minimise_merged_pdbs(system, '_novo')
gen.file_copy_and_check('merged_cg2at_novo_minimised.pdb',
g_var.final_dir + 'final_cg2at_de_novo.pdb')
time_counter['m_t'] = time.time()
#### copies all itp files and topologies from whereever they are stored
for file_name in os.listdir(g_var.working_dir + 'MERGED'):
if file_name.endswith('.itp') or file_name.endswith('final.top'):
gen.file_copy_and_check(
g_var.working_dir + 'MERGED/' + file_name,
g_var.final_dir + file_name)
if 'PROTEIN' in cg_residues:
#### creates mdp file if user wants to pull the structure to initial input
if not os.path.exists(g_var.final_dir + 'steered_md.mdp'):
with open(g_var.final_dir + 'steered_md.mdp', 'w') as steered_md:
steered_md.write(
'define = -DPOSRES\nintegrator = md\nnsteps = 3000\ndt = 0.001\ncontinuation = no\nconstraint_algorithm = lincs\n'
)
steered_md.write(
'constraints = h-bonds\nns_type = grid\nnstlist = 25\nrlist = 1\nrcoulomb = 1\nrvdw = 1\ncoulombtype = PME\n'
)
steered_md.write(
'pme_order = 4\nfourierspacing = 0.16\ntcoupl = V-rescale\ntc-grps = system\ntau_t = 0.1\nref_t = 310\npcoupl = no\n'
)
steered_md.write(
'pbc = xyz\nDispCorr = no\ngen_vel = yes\ngen_temp = 310\ngen_seed = -1'
)
#### calculates final RMS
RMSD = {}
de_novo_atoms = at_mod_p.read_in_atomistic(
g_var.final_dir + 'final_cg2at_de_novo.pdb', system['PROTEIN'],
sequence, False)
RMSD['de novo '] = at_mod_p.RMSD_measure(de_novo_atoms, system,
backbone_coords)
if user_at_input and 'PROTEIN' in system:
at_input_atoms = at_mod_p.read_in_atomistic(
g_var.final_dir + 'final_cg2at_at_rep_user_supplied.pdb',
system['PROTEIN'], sequence, False)
RMSD['at input'] = at_mod_p.RMSD_measure(at_input_atoms, system,
backbone_coords)
print('\n{0:^10}{1:^25}{2:^10}'.format('output ', 'chain',
'RMSD (' + chr(197) + ')'))
print('{0:^10}{1:^25}{2:^10}'.format('-------', '-----', '---------'))
for rmsd in RMSD:
for chain in RMSD[rmsd]:
print('{0:^10}{1:^25}{2:^10}'.format(rmsd, str(chain),
float(RMSD[rmsd][chain])))
#### removes temp file from script, anything with temp in really
if g_var.clean:
gen.clean(cg_residues)
time_counter['f_t'] = time.time()
#### prints out system information
print('\n{:-<100}'.format(''))
print('{0:^100}'.format('Script has completed, time for a beer'))
print('\n{0:^20}{1:^10}'.format('molecules', 'number'))
print('{0:^20}{1:^10}'.format('---------', '------'))
for section in system:
print('{0:^20}{1:^10}'.format(section, system[section]))
#### prints out script timings for each section
if g_var.v >= 1:
gen.print_script_timings(time_counter, system, user_at_input)