def perform_screen(work_name, user_target, center_x, center_y, center_z,
size_x, size_y, size_z, mol_db, pdb_path, res_path):
"""
用户指定筛选中心以及盒子大小
:param work_name:
:param center_x:
:param center_y:
:param center_z:
:param size_x:
:param size_y:
:param size_z:
:param mol_db:
:param pdb_path:
:param res_path:
:return:
"""
screen_status(work_name, status='computing')
email = UserProfile.objects.filter(id=VirtualScreen.objects.filter(
pdb_file=pdb_path.replace('/home/wz/pywork/try27/drug/media', '')
).values()[0]['user_id']).values()[0]['email']
screen_out = 'screen_out.csv'
res = os.path.join(res_path, 'res')
if not os.path.exists(res):
os.mkdir(res)
ligand_db = os.path.join(drugdb, mol_db)
os.system("python %s/extra_apps/vina/prepare_receptor4.py -r %s"
" -A checkhydrogens " % (BASE_DIR, pdb_path))
pdbqt = pdb_path.split('/')[-1].split('.')[0] + '.pdbqt'
if os.path.exists(pdbqt):
os.system("mv %s %s" % (pdbqt, res_path))
if len(user_target) > 1:
user_target = user_target.split(';')
target_list = os.listdir(os.path.join(TARGET_FOLDER_BASE, 'maccs'))
smile_file = os.path.join(ligand_db, 'smiles.csv')
df = pd.read_csv(smile_file, header=None, encoding='utf-8')
smile_data = df.values.tolist()
curr_proc = mp.current_process()
curr_proc.daemon = False
p = mp.Pool(processes=mp.cpu_count())
curr_proc.daemon = True
pool_lst = []
for ligand in smile_data:
smiles = ligand[1]
targets = p.apply_async(pred, args=(smiles, target_list))
pool_lst.append([ligand[0], targets])
p.close()
p.join()
pool_lst = [[n[0], n[1].get()] for n in pool_lst]
for target in pool_lst:
if target[1]:
pred_target = []
for pred_tar in target[1]:
pred_target.append(pred_tar['chembl_id'])
same_target = [l for l in pred_target if l in user_target]
if same_target:
ligand = target[0].split('.')[0] + '.pdbqt'
ligand_path = os.path.join(ligand_db, ligand)
os.system(
"%s --receptor %s/%s --ligand %s --center_x %s --center_y %s"
" --center_z %s --size_x %s --size_y %s --size_z %s" %
('vina', res_path, pdbqt, ligand_path, center_x,
center_y, center_z, size_x, size_y, size_z))
os.system("mv %s %s" %
(ligand_path.split('.')[0] + '_out.pdbqt', res))
os.system(
"python %s/extra_apps/vina/pdbqt_to_pdb.py -f %s -v" %
(BASE_DIR,
os.path.join(
res,
ligand_path.split('/')[-1].split('.')[0] +
'_out.pdbqt')))
else:
ligand_lst = os.listdir(ligand_db)
ligand_lst = [v for v in ligand_lst if v.endswith('pdbqt')]
for ligand_ in ligand_lst:
ligand_path = os.path.join(ligand_db, ligand_)
os.system(
"%s --receptor %s/%s --ligand %s --center_x %s --center_y %s"
" --center_z %s --size_x %s --size_y %s --size_z %s" %
('vina', res_path, pdbqt, ligand_path, center_x, center_y,
center_z, size_x, size_y, size_z))
os.system("mv %s %s" %
(ligand_path.split('.')[0] + '_out.pdbqt', res))
os.system(
"python %s/extra_apps/vina/pdbqt_to_pdb.py -f %s -v" %
(BASE_DIR,
os.path.join(
res,
ligand_path.split('/')[-1].split('.')[0] + '_out.pdbqt')))
res_lst = os.listdir(res)
reg = 'REMARK VINA RESULT:(.*?)\n'
re_reg = re.compile(reg)
screen_res = []
insert_lst = []
for out in res_lst[:]:
out_path = os.path.join(res, out)
with open(out_path, 'r') as f:
data = f.read()
out_lst = re_reg.findall(data)
if out_lst:
med = []
for model in out_lst:
med.append(float(model.split()[0]))
file_name = out.split('_out')[0]
screen_res.append([file_name, min(med)])
insert_lst.append(
Screen(work_name=work_name,
screen_cat='screen',
affinity=min(med),
path=os.path.join(res.split('media/')[1], out[:-2])))
if screen_res:
Screen.objects.bulk_create(insert_lst)
arr = np.array(screen_res)
df = pd.DataFrame(arr, columns=['id', 'Affinity (kcal/mol)'])
df = df.sort_values("Affinity (kcal/mol)", ascending=False)
df.to_csv(screen_out, index=False)
os.system("mv %s %s" % (screen_out, res))
email_status(email, res_path + '/res/screen_out.csv')
else:
email_status(email, '')
screen_status(work_name=work_name, status='completed')
def perform_screen2_user(work_name, user_target, user_db_name, pdb_path,
resi_path, res_path):
"""
用户提供数据库以及残基进行筛选
:param work_name:
:param user_db_name:
:param pdb_path:
:param resi_path:
:param res_path:
:return:
"""
# screen2_status(work_name, status='computing')
with open(resi_path, 'r') as f:
lines = f.readlines()
lines = [n.rstrip() for n in lines if len(n) > 1]
x, y, z = [], [], []
for n in lines:
x.append(float(n[30:38]))
y.append(float(n[38:46]))
z.append(float(n[46:54]))
center_x = float('%.3f' % (sum(x) / len(x)))
center_y = float('%.3f' % (sum(y) / len(y)))
center_z = float('%.3f' % (sum(z) / len(z)))
size_x = max(x) - min(x)
size_y = max(y) - min(y)
size_z = max(z) - min(z)
screen_out = 'screen_out.csv'
user_db = os.path.join(res_path, 'userdb')
res = os.path.join(res_path, 'res')
if not os.path.exists(user_db):
os.mkdir(user_db)
if not os.path.exists(res):
os.mkdir(res)
input_file = os.path.join(res_path, user_db_name)
gen_user_db_qt_smiles(input_file, user_db)
os.system("python %s/extra_apps/vina/prepare_receptor4.py -r /%s"
" -A checkhydrogens " % (BASE_DIR, pdb_path))
pdbqt = pdb_path.split('/')[-1].split('.')[0] + '.pdbqt'
if os.path.exists(pdbqt):
os.system("mv %s %s" % (pdbqt, res_path))
if len(user_target) > 1:
user_target = user_target.split(';')
target_list = os.listdir(os.path.join(TARGET_FOLDER_BASE, 'maccs'))
smile_file = os.path.join(user_db, 'smiles.csv')
df = pd.read_csv(smile_file, header=None, encoding='utf-8')
smile_data = df.values.tolist()
curr_proc = mp.current_process()
curr_proc.daemon = False
p = mp.Pool(processes=mp.cpu_count())
curr_proc.daemon = True
pool_lst = []
for ligand in smile_data:
smiles = ligand[1]
targets = p.apply_async(pred, args=(smiles, target_list))
pool_lst.append([ligand[0], targets])
p.close()
p.join()
pool_lst = [[n[0], n[1].get()] for n in pool_lst]
for target in pool_lst:
if target[1]:
pred_target = []
for pred_tar in target[1]:
pred_target.append(pred_tar['chembl_id'])
same_target = [l for l in pred_target if l in user_target]
if same_target:
ligand = target[0].split('.')[0] + '.pdbqt'
ligand_path = os.path.join(user_db, ligand)
if os.path.exists(ligand_path):
os.system(
"%s --receptor %s/%s --ligand %s --center_x %s --center_y %s"
" --center_z %s --size_x %s --size_y %s --size_z %s"
% ('vina', res_path, pdbqt, ligand_path, center_x,
center_y, center_z, size_x, size_y, size_z))
os.system(
"mv %s %s" %
(ligand_path.split('.')[0] + '_out.pdbqt', res))
os.system(
"python %s/extra_apps/vina/pdbqt_to_pdb.py -f %s -v"
% (BASE_DIR,
os.path.join(
res,
ligand_path.split('/')[-1].split('.')[0] +
'_out.pdbqt')))
else:
ligand_lst = os.listdir(user_db)
ligand_lst = [v for v in ligand_lst if v.endswith('pdbqt')]
for ligand_ in ligand_lst:
ligand_path = os.path.join(user_db, ligand_)
os.system(
"%s --receptor %s/%s --ligand %s --center_x %s --center_y %s"
" --center_z %s --size_x %s --size_y %s --size_z %s" %
('vina', res_path, pdbqt, ligand_path, center_x, center_y,
center_z, size_x, size_y, size_z))
os.system("mv %s %s" %
(ligand_path.split('.')[0] + '_out.pdbqt', res))
os.system(
"python %s/extra_apps/vina/pdbqt_to_pdb.py -f %s -v" %
(BASE_DIR,
os.path.join(
res,
ligand_path.split('/')[-1].split('.')[0] + '_out.pdbqt')))
res_lst = os.listdir(res)
reg = 'REMARK VINA RESULT:(.*?)\n'
re_reg = re.compile(reg)
screen_res = []
insert_lst = []
for out in res_lst[:]:
out_path = os.path.join(res, out)
with open(out_path, 'r') as f:
data = f.read()
out_lst = re_reg.findall(data)
if out_lst:
med = []
for model in out_lst:
med.append(float(model.split()[0]))
file_name = out.split('_out')[0]
screen_res.append([file_name, min(med)])
insert_lst.append(
Screen(work_name=work_name,
screen_cat='screen2',
affinity=min(med),
path=os.path.join(res.split('media/')[1], out[:-2])))
if screen_res:
Screen.objects.bulk_create(insert_lst)
arr = np.array(screen_res)
df = pd.DataFrame(arr, columns=['id', 'Affinity (kcal/mol)'])
df = df.sort_values("Affinity (kcal/mol)", ascending=False)
df.to_csv(screen_out, index=False)
os.system("mv %s %s" % (screen_out, res))
def perform_screen_user(work_name, center_x, center_y, center_z, size_x,
size_y, size_z, user_db_name, pdb_path, res_path):
"""
用户提供数据库以及中心坐标和盒子大小进行筛选
:param work_name:
:param center_x:
:param center_y:
:param center_z:
:param size_x:
:param size_y:
:param size_z:
:param user_db_name:
:param pdb_path:
:param res_path:
:return:
"""
screen_status(work_name, status='computing')
screen_out = 'screen_out.csv'
user_db = os.path.join(res_path, 'userdb')
res = os.path.join(res_path, 'res')
if not os.path.exists(user_db):
os.mkdir(user_db)
if not os.path.exists(res):
os.mkdir(res)
input_file = os.path.join(res_path, user_db_name)
gen_user_db_qt_smiles(input_file, user_db)
os.system("python %s/extra_apps/vina/prepare_receptor4.py -r /%s"
" -A checkhydrogens " % (BASE_DIR, pdb_path))
pdbqt = pdb_path.split('/')[-1].split('.')[0] + '.pdbqt'
os.system("mv %s %s" % (pdbqt, res_path))
target_list = os.listdir(os.path.join(TARGET_FOLDER_BASE, 'maccs'))
smile_file = os.path.join(user_db, 'smiles.csv')
df = pd.read_csv(smile_file, header=None, encoding='utf-8')
smile_data = df.values.tolist()
curr_proc = mp.current_process()
curr_proc.daemon = False
p = mp.Pool(processes=mp.cpu_count())
curr_proc.daemon = True
pool_lst = []
for ligand in smile_data:
smiles = ligand[1]
targets = p.apply_async(pred, args=(smiles, target_list))
pool_lst.append([ligand[0], targets])
p.close()
p.join()
pool_lst = [[n[0], n[1].get()] for n in pool_lst]
for target in pool_lst:
if target[1]:
ligand = target[0].split('.')[0] + '.pdbqt'
ligand_path = os.path.join(user_db, ligand)
if os.path.exists(ligand_path):
os.system(
"%s --receptor %s/%s --ligand %s --center_x %s --center_y %s"
" --center_z %s --size_x %s --size_y %s --size_z %s" %
(vina_path, res_path, pdbqt, ligand_path, center_x,
center_y, center_z, size_x, size_y, size_z))
os.system("mv %s %s" %
(ligand_path.split('.')[0] + '_out.pdbqt', res))
os.system(
"python %s/extra_apps/vina/pdbqt_to_pdb.py -f %s -v" %
(BASE_DIR,
os.path.join(
res,
ligand_path.split('/')[-1].split('.')[0] +
'_out.pdbqt')))
res_lst = os.listdir(res)
reg = 'REMARK VINA RESULT:(.*?)\n'
re_reg = re.compile(reg)
screen_res = []
insert_lst = []
for out in res_lst[:]:
out_path = os.path.join(res, out)
with open(out_path, 'r') as f:
data = f.read()
out_lst = re_reg.findall(data)
if out_lst:
med = []
for model in out_lst:
med.append(float(model.split()[0]))
file_name = out.split('_out')[0]
screen_res.append([file_name, min(med)])
insert_lst.append(
Screen(work_name=work_name,
screen_cat='screen',
affinity=min(med),
path=os.path.join(res.split('media/')[1], out[:-2])))
Screen.objects.bulk_create(insert_lst)
arr = np.array(screen_res)
df = pd.DataFrame(arr, columns=['id', 'Affinity (kcal/mol)'])
df = df.sort_values("Affinity (kcal/mol)", ascending=False)
df.to_csv(screen_out, index=False)
os.system("mv %s %s" % (screen_out, res))
screen_status(work_name, status='completed')