def calculate_single(pdb_id, cid, seq_marker):
seq_marker = int(seq_marker)
pdb_data = pdb_interface.get_info_for(pdb_id)
if pdb_data is None:
raise Exception(f"PDB file download fail for {pdb_id}.")
ATOM, HETATM, hb = pdb_data
try:
dsr_snos = _get_sno_range(ATOM, cid, seq_marker)
if dsr_snos is None or len(dsr_snos) != 30 or dsr_snos[0] != seq_marker:
msg = f"ATOM lines not found in range({seq_marker}, "
f"{seq_marker + 30}) for {pdb_id}:{cid}.<br>"
raise Exception(msg)
res, C, CA, N = _from_considered_elements_single(ATOM, dsr_snos, cid)
pept_bonds = _get_pept_bonds(CA, dsr_snos)
# For filling descr df
res_CA = _get_res_CA(res, CA, dsr_snos)
angles, CA = dihedrals.get_descr_dihedrals(C, CA, N, dsr_snos)
hbond_descr = hbonds.get_descr_hb(hb, ATOM, HETATM, dsr_snos)
heavy_atom_contacts, hetatom_contacts, hetatom_covalent = \
contacts.get_contacts(ATOM, HETATM, cid, dsr_snos)
descr = _assemble_descr(hetatom_contacts, hetatom_covalent,
heavy_atom_contacts, angles, hbond_descr,
res_CA, pept_bonds)
full_descr = _add_columns(descr, pdb_id, seq_marker, cid)
except Exception as e:
msg = f"Exception caught in descriptor calculation. Traceback: " \
f"<{traceback.format_exc()}>. Error: <{e}>"
raise Exception(msg)
return full_descr
def find(matrix_file, num_seqs, pdb_seq_file, output, motif_len=30):
conv_output = paths.CONV_OUTPUT
conv_interface.run(matrix_file, motif_len, num_seqs, conv_output)
pdb_cid_motif_raw = search_converters.search_run(conv_output, pdb_seq_file)
pdb_cids = []
for pdb_id, values in pdb_cid_motif_raw.items():
for cid in values['cid']:
pdb_cids.append((pdb_id, cid))
pdb_cid_seq = dict()
print(len(pdb_cids))
if os.path.isfile(paths.RCSB_SEQS):
with open(paths.RCSB_SEQS, 'rb') as file:
rcsb_seqs = pickle.load(file)
else:
rcsb_seqs = dict()
for i, (pdb_id, cid) in enumerate(pdb_cids):
if not i % 10:
print(i)
if (pdb_id.upper(), cid.upper()) in rcsb_seqs:
pdb_cid_seq[(pdb_id, cid)] = rcsb_seqs[(pdb_id.upper(),
cid.upper())]
else:
try:
ATOM = pdb_interface.get_info_for(pdb_id)[0]
ATOM_cid = ATOM[ATOM.cid == cid]
if ATOM_cid is None:
continue
seq = pdb_interface._extract_seq_from_df(ATOM_cid)
if seq is None:
continue
except Exception as e:
print(f"get_seq_for() fails for pdb_id/cid {pdb_id}/{cid}. "
f"Skipping.")
print(f"Traceback: <{traceback.format_exc()}>")
print(f"Error_msg: <{e}>")
continue
pdb_cid_seq[(pdb_id, cid)] = seq
rcsb_seqs[(pdb_id.upper(), cid.upper())] = seq
with open(paths.RCSB_SEQS, 'wb') as file:
pickle.dump(rcsb_seqs, file, -1)
pdb_cid_seq = OrderedDict(sorted(pdb_cid_seq.items()))
pdb_structure_seqs = os.path.join(paths.DEBUG, "pdb_structure_seqs.txt")
with open(pdb_structure_seqs, 'w') as file:
for (pdb_id, cid), seq in pdb_cid_seq.items():
file.write(f">{pdb_id}_{cid}\n")
file.write(seq + "\n")
clean_fasta_alphabet.screen(pdb_structure_seqs, pdb_structure_seqs)
filter_seqs.delete_short_seqs(pdb_structure_seqs, motif_len)
motif_positions = search_converters.search_run(conv_output,
pdb_structure_seqs)
with open(output, 'wb') as file:
pickle.dump(motif_positions, file, -1)
def calculate(motif_pos_map):
descrs = pd.DataFrame()
print(f"Total length: {len(motif_pos_map)}.")
print(len(motif_pos_map))
for i, (pdb_id, motif_cid_map) in enumerate(motif_pos_map.items()):
if not (i % 10):
print(i)
print(f"{len(motif_pos_map) - i}: {pdb_id}")
motif_pos_s = motif_cid_map['sno_markers']
cids = motif_cid_map['cid']
pdb_data = pdb_interface.get_info_for(pdb_id)
if pdb_data is None:
continue
ATOM, HETATM, hb = pdb_data
if not isinstance(motif_pos_s, list):
motif_pos_s = [motif_pos_s]
cids = [cids]
for motif_pos, cid in zip(motif_pos_s, cids):
try:
dsr_snos = _get_sno_range(ATOM, cid, motif_pos)
if dsr_snos is None:
continue
res, C, CA, N = _from_considered_elements(ATOM, dsr_snos, cid)
pept_bonds = _get_pept_bonds(CA, dsr_snos)
# For filling descr df
res_CA = _get_res_CA(res, CA, dsr_snos)
angles, CA = dihedrals.get_descr_dihedrals(C, CA, N, dsr_snos)
hbond_descr = hbonds.get_descr_hb(hb, ATOM, HETATM, dsr_snos)
heavy_atom_contacts, hetatom_contacts, hetatom_covalent = \
contacts.get_contacts(ATOM, HETATM, cid, dsr_snos)
descr = _assemble_descr(hetatom_contacts, hetatom_covalent,
heavy_atom_contacts, angles, hbond_descr,
res_CA, pept_bonds)
full_descr = _add_columns(descr, pdb_id, motif_pos, cid)
descrs = descrs.append(full_descr, ignore_index=True)
except Exception as e:
print(e)
print(f"Calc_descr failed for {pdb_id}:{cid}")
pdb_suffix = pdb_id.lower().strip()
if pdb_suffix+".pkl" in paths.PDB_PARSED_SET:
os.remove(os.path.join(paths.PDB_PARSED,
pdb_suffix + ".pkl"))
# raise
continue
return descrs
def calculate(motif_pos_map):
descrs = pd.DataFrame()
i = 0
for pdb_id, motif_cid_map in motif_pos_map.items():
i += 1
# if pdb_id != "2xsx":
# continue
motif_pos = motif_cid_map['sno_markers']
cid = motif_cid_map['cid']
pdb_data = pdb_interface.get_info_for(pdb_id)
if pdb_data is None:
continue
ATOM, HETATM, hb = pdb_data
try:
motif_pos = motif_pos[0]
dsr_snos = _get_sno_range(ATOM, cid, motif_pos)
if dsr_snos is None:
continue
res, C, CA, N = _from_considered_elements(ATOM, dsr_snos, cid)
pept_bonds = _get_pept_bonds(CA, dsr_snos)
# For filling descr df
res_CA = _get_res_CA(res, CA, dsr_snos)
angles, CA = dihedrals.get_descr_dihedrals(C, CA, N, dsr_snos)
hbond_descr = hbonds.get_descr_hb(hb, ATOM, HETATM, dsr_snos)
heavy_atom_contacts, hetatom_contacts, hetatom_covalent = \
contacts.get_contacts(ATOM, HETATM, cid, dsr_snos)
descr = _assemble_descr(hetatom_contacts, hetatom_covalent,
heavy_atom_contacts, angles, hbond_descr,
res_CA, pept_bonds)
full_descr = _add_columns(descr, pdb_id, motif_pos, cid)
descrs = descrs.append(full_descr, ignore_index=True)
except:
print(f"Calc_descr failed for {pdb_id}:{cid}")
raise
continue
return descrs