def _load_structures(inp, sys_params,):
'''load protein/ligand structures needed for calculation'''
parser = pdb.Big_PDBParser()
print("now loading structures")
# Read and/or create pickle files for the structures to save I/O time
ligand_pkl_filename = os.path.join(inp['rootdir'], "ligand.pkl")
receptor_pkl_filename = os.path.join(inp['rootdir'], "receptor.pkl")
receptor_pkl_dry_filename = os.path.join(inp['rootdir'], "receptor_dry.pkl")
receptor_pkl_dry_pqr_filename = os.path.join(inp['rootdir'], "receptor_dry_pqr.pkl")
ligand=pickle_or_load(sys_params['lig_pqr_filename'], ligand_pkl_filename, struc_name="ligand", pqr=True)
#receptor=pickle_or_load(sys_params['rec_pdb_filename'], receptor_pkl_filename, struc_name="receptor", pqr=False)
receptor_dry=pickle_or_load(sys_params['rec_dry_pdb_filename'], receptor_pkl_dry_filename, struc_name="receptor_dry", pqr=False)
receptor_dry_pqr=pickle_or_load(sys_params['rec_dry_pqr_filename'], receptor_pkl_dry_pqr_filename, struc_name="receptor_dry_pqr", pqr=True)
struct={ # all parameters pertaining to structure
'ligand':ligand,
#'receptor':receptor,
'receptor_dry':receptor_dry, # or create a function that will remove all waters, complicated by ions
'receptor_dry_pqr':receptor_dry_pqr,
'rec_com':pdb.center_of_mass(receptor_dry), # have to take into account the center of mass of the receptor itself
'lig_com':pdb.center_of_mass(ligand), # have to take into account the center of mass of the ligand itself
}
return struct
def _parse_pdb(
filename,
struct_name="pickle",
):
parser = pdb.Big_PDBParser()
struct = parser.get_structure(struct_name,
filename,
pqr=False,
conventional=False) # load the file
return struct
def pickle_or_load(filename, picklename, struc_name="pickle",pqr=False):
'''for large files, instead of parsing, they can be saved and loaded much more quickly as a pickle. '''
parser = pdb.Big_PDBParser()
if os.path.exists(picklename) and os.path.getmtime(picklename) > os.path.getmtime(filename): # if the pickle has been most recently modified
# load the pickle
print("reading pickle:", picklename)
our_file=open(picklename, 'rb')
our_obj=pickle.load(our_file)
our_file.close()
else:
# then load the file itself and save the pickle
our_obj=parser.get_structure(struc_name, filename, pqr=pqr, conventional=False) # load the file
print("writing pickle:", picklename)
our_file=open(picklename, 'wb')
pickle.dump(our_obj, our_file, protocol=-1)
our_file.close()
return our_obj
import milestones
import numpy as np
import pdb2 as pdb
from copy import deepcopy # needed to keep track of separate structure objects
import namd_inputs # contains all the details for the namd input files
import colvars
import psfmerge
#import MDAnalysis as mda #import *
from adv_template import Adv_template, File_template
import unittest
import positions_orient
import cPickle as pickle
verbose = True
parser = pdb.Big_PDBParser()
self_path = os.path.dirname(
os.path.realpath(__file__)) # get the path to this script
fwd_rev_template_location = os.path.join(self_path, 'fwd_rev.template')
def amber_building(settings):
'''the pre-minimization procedure for amber ff simulations'''
i = settings['index']
building = settings['md_file_paths'][i]['building']
working_pdb_base = 'holo' #settings['working_pdb_base']
pdbfile = settings['md_file_paths'][i]['wet_holo']
amber_settings = settings['amber_settings']
#sources=amber_settings['sources']
def build_bd(seekrcalc):
'''
build all structures and necessary files for BD calculations
takes a list of pqr filenames, reaction coordinate pairs/distances
'''
if verbose: print '\n', '#'*40, "\n Now creating BD files using bd.py\n", '#'*40
parser = pdb.Big_PDBParser()
rec_struct = parser.get_structure('bd_receptor_dry_pqr', seekrcalc.browndye.rec_dry_pqr_filename, pqr=True)
#milestone_pos_rot_list = settings['milestone_pos_rot_list'] # NOTE: may want to clean up code referring to this variable
''' # TODO: marked for removal because feature probably not needed
if settings['starting_conditions'] == 'configs':
lig_configs = settings['lig_configs']
elif settings['starting_conditions'] == 'spheres':
# then generate the anchors from random positions/orientations in a sphere
# NOTE: there is a random orientation function in positions_orient.py
pass
else:
raise Exception, "option not allowed: %s" % settings['starting_conditions']
'''
#bd_file_paths = settings['bd_file_paths']
#browndye_bin = settings['browndye_bin_dir']
#if not os.path.exists(empty_pqrxml)
bd_configs = []
# fill the b_surface folder
lig_config = seekrcalc.browndye.starting_lig_config # get the first config of the ligand
lig_center = pdb.center_of_mass(lig_config)
pqrs = [copy.deepcopy(rec_struct), copy.deepcopy(lig_config)]
pqrs[1].struct_id='bd_ligand'
#for site in settings['b_surface_ending_surfaces']:
b_surface_criteria = []
starting_surfaces = []
for milestone in seekrcalc.milestones:
if milestone.end:
b_surface_criteria.append({'centerx':milestone.center_vec[0], 'centery':milestone.center_vec[1], 'centerz':milestone.center_vec[2], 'ligx':lig_center[0], 'ligy':lig_center[1], 'ligz':lig_center[2], 'radius':milestone.radius, 'index':milestone.index, 'siteid':milestone.siteid}) # add every site to the criteria list
if milestone.bd:
starting_surfaces.append({'site':milestone.siteid, 'radius':milestone.radius, 'index':milestone.index})
print "bsurface_criteria:", b_surface_criteria
b_surface_pqrxmls = write_browndye_input(pqrs, seekrcalc, b_surface_criteria, work_dir=seekrcalc.browndye.b_surface_path, browndye_bin=seekrcalc.browndye.browndye_bin, start_at_site='false', fhpd_mode = False) # write input for this part
'''
for bd_file_path in bd_file_paths:
if not bd_file_path:
pass
#bd_configs.append(None)
#continue # then don't do BD for this portion
#print "bd_file_path:", bd_file_path
anchor_folder_name = bd_file_path.split('/')[-2] # reading the file tree to get the index of every existing folder
#print "anchor_folder_name:", anchor_folder_name
folder_index = anchor_folder_name.split('_')[1] # getting the index out of the folder name
bd_configs.append(int(folder_index))
'''
counter = 0 # the index of the loop itself
for milestone in seekrcalc.milestones:
if milestone.bd:
lig_config = milestone.config
bd_file_path = os.path.join(seekrcalc.project.rootdir, milestone.directory, 'bd')
print "bd_file_path:", bd_file_path
pqrs = [copy.deepcopy(rec_struct), copy.deepcopy(lig_config)]
pqrs[1].struct_id='bd_ligand'
bd_needed = True
'''
m=0
for m in range(len(milestone_pos_rot_list)): # m will represent the proper milestone index
if milestone_pos_rot_list[m][0].index == i:
break
'''
criteria = []
for surface in starting_surfaces:
if surface['site'] == milestone.siteid:
proper_radius = milestone.bd_adjacent.radius
proper_index = milestone.bd_adjacent.index
else:
proper_radius = surface['radius']
proper_index = surface['index']
criteria.append({'centerx':milestone.center_vec[0], 'centery':milestone.center_vec[1], 'centerz':milestone.center_vec[2], 'ligx':lig_center[0], 'ligy':lig_center[1], 'ligz':lig_center[2], 'radius':proper_radius, 'index':proper_index, 'siteid':milestone.siteid})
'''
criteria = [] #[[(31.121, 37.153, 35.253), (38.742, 51.710, 68.137), 9.0],] # a list of all reaction criteria
for site in settings['starting_surfaces']:
#site_center = [site['x'], site['y'],site['z']]
#radius = site['radius'] # NOTE: at this time, only spherical reaction criteria are allowed in BrownDye
if milestone_pos_rot_list[m][0].siteid == site['siteid']: # then its the same site, we need to go one shell in
proper_radius = site['inner_radius'] # the radius in the same site
proper_index = site['inner_index']
else: # this is a different site, choose the same radius as starting
proper_radius = site['outer_radius']
proper_index = site['outer_index']
criteria.append({'centerx':site['x'], 'centery':site['y'], 'centerz':site['z'], 'ligx':lig_center[0], 'ligy':lig_center[1], 'ligz':lig_center[2], 'radius':proper_radius, 'index':proper_index, 'siteid':site['siteid']}) # add every site to the criteria list
'''
#print "pqrs:", pqrs, 'criteria:', criteria
site_pqrxmls = write_browndye_input(pqrs, seekrcalc, criteria, work_dir=bd_file_path, browndye_bin=seekrcalc.browndye.browndye_bin,fhpd_mode=True)
# make BD preparation scripts extract_bd_frames.py and bd_fhpd.pyp
# Write the script to extract all frames from the successful b_surface bd trajectories
extract_bd_frames_dict = {'TRAJDIR':"../../b_surface", 'WORKDIR':"./trajs", 'PQRXML0':os.path.basename(b_surface_pqrxmls[0]), 'PQRXML1':os.path.basename(b_surface_pqrxmls[1]), 'EMPTY':empty_pqrxml, 'SITENAME':'%s_%s' % (milestone.siteid, milestone.index), 'NUMBER_OF_TRAJS':seekrcalc.browndye.num_threads}
extract_bd_frames = Adv_template(extract_bd_frames_template, extract_bd_frames_dict)
extract_file = open(os.path.join(bd_file_path,"extract_bd_frames.py"), 'w')
extract_file.write(extract_bd_frames.get_output()) # write an xml file for the input to bd
extract_file.close()
# construct the FHPD distribution prep scripts
make_fhpd_dict = {'INPUT_TEMPLATE_FILENAME':'input.xml', 'RECEPTOR_PQRXML':os.path.basename(b_surface_pqrxmls[0]), 'RXNS':'rxns.xml', 'NTRAJ':seekrcalc.browndye.fhpd_numtraj, 'ARGS':"glob.glob(os.path.join('./trajs','lig*.pqr'))"} # NOTE: should change NTRAJ to be consistent with the number of reaction events in the b_surface phase
make_fhpd = Adv_template(make_fhpd_template,make_fhpd_dict)
make_fhpd_file = open(os.path.join(bd_file_path,"make_fhpd.py"), 'w')
make_fhpd_file.write(make_fhpd.get_output()) # write an xml file for the input to bd
make_fhpd_file.close()
# Consolidate all result files from the FHPD simulations into one large results.xml file
fhpd_consolidate_dict = {'FHPD_DIR':"fhpd", 'LIG_DIR_GLOB':"lig*/", 'RESULTS_NAME':'results.xml'}
fhpd_consolidate = Adv_template(fhpd_consolidate_template,fhpd_consolidate_dict)
fhpd_consolidate_file = open(os.path.join(bd_file_path,"fhpd_consolidate.py"), 'w')
fhpd_consolidate_file.write(fhpd_consolidate.get_output()) # write an xml file for the input to bd
fhpd_consolidate_file.close()
make_empty_pqrxml(os.path.join(bd_file_path, 'empty.pqrxml'))
counter += 1
def generate_configs(seekrcalc):
'''Generates the apo and holo configurations of the receptors and ligands.
Input:
- seekrcalc: a SeekrCalculation() object that contains all the settings for
the SEEKR calculation
Output:
- None
'''
if verbose: print "Generating structural configurations..."
parser = pdb.Big_PDBParser()
if verbose: print "now loading structures"
# Read and/or create pickle files for the structures to save I/O time
ligand_pkl_filename = os.path.join(seekrcalc.project.rootdir, "ligand.pkl")
receptor_pkl_wet_filename = os.path.join(seekrcalc.project.rootdir,
"receptor.pkl")
#receptor_pkl_dry_filename = os.path.join(seekrcalc.project.rootdir, "receptor_dry.pkl")
receptor_pkl_dry_pqr_filename = os.path.join(seekrcalc.project.rootdir,
"receptor_dry_pqr.pkl")
ligand = pickle_or_load(seekrcalc.building.lig_dry_pqr_filename,
ligand_pkl_filename,
parser,
struc_name="ligand",
pqr=True)
receptor_wet = pickle_or_load(seekrcalc.building.rec_wet_pdb_filename,
receptor_pkl_wet_filename,
parser,
struc_name="receptor_wet",
pqr=False)
receptor_dry = pickle_or_load(seekrcalc.building.rec_dry_pqr_filename,
receptor_pkl_dry_pqr_filename,
parser,
struc_name="receptor_dry_pqr",
pqr=True)
seekrcalc.building.ligand = ligand
seekrcalc.building.rec_wet_pdb_filename = receptor_wet
seekrcalc.building.rec_dry_pqr_filename = receptor_dry
milestones = seekrcalc.milestones
configs = []
starttime = time.time()
struct_center = pdb.center_of_mass(ligand)
if verbose: print "Generating", len(milestones), "ligand configurations"
for milestone in milestones:
new_ligand = deepcopy(ligand) # copy the entire structure
new_ligand.moveby(
-struct_center) # reposition ligand structure over the origin
new_ligand.moveby(
milestone.anchor) # move ligand to the anchor location
new_ligand.struct_id = milestone.fullname
configs.append(new_ligand)
endtime = time.time()
if verbose:
"Generate_configs complete. Total number: %d. Elapsed time: %d s" % (
len(configs),
endtime - starttime,
)
print "Now running through all configurations to combine ligand with receptor."
if seekrcalc.building.reject_clashes:
print "Rejecting steric clashes..."
else:
print "Ignoring steric clashes..."
if seekrcalc.project.bd:
seekrcalc.browndye.starting_lig_config = configs[0]
for i, milestone in enumerate(
milestones): # run thru all the configurations of ligands
lig_config = configs[i]
milestone.config = deepcopy(lig_config)
if seekrcalc.building.reject_clashes:
if structures_clash(lig_config, receptor_dry, tolerance=0.2):
continue # if there's a clash
if milestone.md:
#pdb.TER_resnames.append(seekrcalc.building.lig_resname) # TODO: marked for removal
holo_config_wet, insert_index, last_ligand_index = pdb.ligmerge(
lig_config, receptor_wet, verbose=False)
holo_config_wet.struct_id = lig_config.struct_id # set the structure description to the same as the ligand
holo_config_wet.renumber_indeces(
) # to number the indeces consecutively
wet_holo_filename = milestone.openmm.wet_holo_pdb_filename
if verbose: print "writing file:", wet_holo_filename
holo_config_wet.save(
wet_holo_filename, amber=True, standard=False
) # write the holo structure into the md directory
last_insert_index = insert_index
last_last_ligand_index = last_ligand_index
if milestone.bd:
holo_config_dry, insert_index, last_ligand_index = pdb.ligmerge(
lig_config, receptor_dry, verbose=False)
holo_config_dry.struct_id = lig_config.struct_id # set the structure description to the same as the ligand
holo_config_dry.renumber_indeces(
) # to number the indeces consecutively
dry_holo_filename = milestone.openmm.dry_holo_pdb_filename
if verbose: print "writing file:", dry_holo_filename
holo_config_dry.save(
dry_holo_filename, amber=True, standard=False
) # write the holo structure into the md directory
return holo_config_wet, last_insert_index, last_last_ligand_index