def mk_table_seq(traj_file, topo_file):
"""This function extracts Protein Block sequences from a trajectory file using pbxplore
librairy https://pbxplore.readthedocs.io/en/latest/ and stock the sequences in table_seq
DataFrame.
Parameters:
traj_file: string, path to a trajectory file.
topo_file: string, path to a topology file.
Output:
table_seq: a pandas.DataFrame containing each sequence of PB extracted.
"""
table_seq = pd.DataFrame()
# Counter for frame for row name
i = 0
for chain_name, chain in pbx.chains_from_trajectory(traj_file, topo_file):
i += 1
# Get dihedrals angles to assign a PB to each position
dihedrals = chain.get_phi_psi_angles()
pb_seq = pbx.assign(dihedrals)
# pbxplore need a tresholds of 5 positions to a assign a PB to a positions. Consequently the two first
# and the two last PB of a sequence are Z which undertermined, we remove them.
table_seq = pd.concat(
[table_seq,
pd.DataFrame(list(pb_seq)[2:-2], columns=[i])], axis=1)
# For some reasons, the sequences are assigned by columns, so the table is transposed at the end
# to put them in rows
table_seq = table_seq.transpose()
table_seq.columns = list(range(2, len(table_seq.columns) + 2))
return (table_seq)
def pbassign_cli():
"""
PBassign command line.
"""
options, pdb_name_lst = user_inputs()
if options.p:
if pdb_name_lst:
print("{} PDB file(s) to process".format(len(pdb_name_lst)))
else:
print("Nothing to do. Good bye.")
return
# PB assignement of PDB structures
chains = pbx.chains_from_files(pdb_name_lst)
else:
# PB assignement of a Gromacs trajectory
chains = pbx.chains_from_trajectory(options.x, options.g)
all_comments = []
all_sequences = []
for comment, chain in chains:
dihedrals = chain.get_phi_psi_angles()
sequence = pbx.assign(dihedrals)
all_comments.append(comment)
all_sequences.append(sequence)
fasta_name = options.o + ".PB.fasta"
with open(fasta_name, "w") as outfile:
pbx.io.write_fasta(outfile, all_sequences, all_comments)
print("wrote {0}".format(fasta_name))
def __init__(self, args):
self.md_sa_seq = []
if args.f:
file_output_fasta = open(args.f, "w")
for chain_name, chain in pbx.chains_from_trajectory(\
args.input_trajectory,\
args.input_topology):
dihedrals = chain.get_phi_psi_angles()
pb_seq = pbx.assign(dihedrals)
self.md_sa_seq.append(pb_seq)
if args.f:
file_output_fasta.write(">{}\n".format(chain_name))
file_output_fasta.write("{}\n".format(pb_seq))
if args.f:
file_output_fasta.close()
def test_loader_xtc(self):
"""
Test for API load function on xtc files
"""
topol = os.path.join(here, "test_data/barstar_md_traj.gro")
traj = os.path.join(here, "test_data/barstar_md_traj.xtc")
chains = list(pbx.chains_from_trajectory(traj, topol))
comment, chain = chains[0]
ref_comment = "{0} | frame 0".format(traj)
ref_chain = "Chain / model : 355 atoms"
assert ref_comment == comment
assert ref_chain == format(chain)
comment, chain = chains[-1]
ref_comment = "{0} | frame 9".format(traj)
assert ref_comment == comment
assert ref_chain == format(chain)
def test_loader_xtc(self):
"""
Test for API load function on xtc files
"""
topol = os.path.join(here, "test_data/barstar_md_traj.gro")
traj = os.path.join(here, "test_data/barstar_md_traj.xtc")
chains = list(pbx.chains_from_trajectory(traj, topol))
comment, chain = chains[0]
ref_comment = "{0} | frame 0".format(traj)
ref_chain = "Chain / model : 355 atoms"
self.assertEqual(ref_comment, comment)
self.assertEqual(ref_chain, format(chain))
comment, chain = chains[-1]
ref_comment = "{0} | frame 9".format(traj)
self.assertEqual(ref_comment, comment)
self.assertEqual(ref_chain, format(chain))
def pbassign_func(self):
"""
Assigns a PB sequence (16 Structural Prototypes) to the protein.
The data file is taken from the listbox.
:return:PB Sequences in output fasta file.Call other two
functions to save the resulted fasta file and to count the
probability matrix for PBs.
"""
# Get paths for the trajectory and topology files directly from
# the listbox.
trajectory = self.trajectory_file.get()
topology = self.topology_file.get()
# Start PB assignment
names = []
pb_sequences = []
for chain_name, chain in pbx.chains_from_trajectory(
trajectory=trajectory, topology=topology):
dihedrals = chain.get_phi_psi_angles()
pb_seq = pbx.assign(dihedrals)
names.append(chain_name)
pb_sequences.append(pb_seq)
# Record the progress in the log screen
self.txt_log.insert('end', ">>> PB assignment is finished" + "\n\n")
# Save the Protein Blocks Sequences into one fasta file
# The saving process is done via the function save_fasta()
save_it.save_fasta(input_seqs=pb_sequences,
names=names,
file_name=os.path.basename(trajectory))
# Record the progress in the log screen
self.txt_log.insert(
'end', ">>> PB Sequences were saved into "
"a Fasta file" + "\n\n")
# Call count_occurrence() to compute the occurrence of each PB.
self.count_occurrence(seq=pb_sequences)
def pbassign_cli():
"""
PBassign command line.
"""
options, pdb_name_lst = user_inputs()
if options.p:
if pdb_name_lst:
print("{} PDB file(s) to process".format(len(pdb_name_lst)))
else:
print('Nothing to do. Good bye.')
return
# PB assignement of PDB structures
chains = pbx.chains_from_files(pdb_name_lst)
else:
# PB assignement of a Gromacs trajectory
chains = pbx.chains_from_trajectory(options.x, options.g)
all_comments = []
all_sequences = []
for comment, chain in chains:
try:
dihedrals = chain.get_phi_psi_angles()
sequence = pbx.assign(dihedrals)
all_comments.append(comment)
all_sequences.append(sequence)
except FloatingPointError:
print(
"The computation of angles produced NaN. This typically means there are issues"
" with some residues coordinates. Check your input file ({0})".
format(comment),
file=sys.stderr)
if all_comments:
fasta_name = options.o + ".PB.fasta"
with open(fasta_name, 'w') as outfile:
pbx.io.write_fasta(outfile, all_sequences, all_comments)
print("wrote {0}".format(fasta_name))
else:
print("No output file was written")
def cli(args=None):
"""Entry point for seq_to_first_iso's CLI.
Parameters
----------
args : list of str, optional
CLI arguments, args are used for testing (default is None for CLI).
Returns
-------
None
Writes a csv file and possibly a gml file.
Raises
------
SystemExit
If no sequences were found on the file.
Notes
-----
Main function of the script, for use with CLI.
"""
if not args:
args = sys.argv[1:]
options, pdb_name_lst = user_inputs()
if options.pdb:
if pdb_name_lst:
print("{} PDB file(s) to process".format(len(pdb_name_lst)))
else:
print('Nothing to do. Good bye.')
return
# PB assignement of PDB structures
chains = pbx.chains_from_files(pdb_name_lst)
else:
# PB assignement of a Gromacs trajectory
chains = pbx.chains_from_trajectory(options.x, options.g)
all_comments = []
all_sequences = []
for comment, chain in chains:
try:
dihedrals = chain.get_phi_psi_angles()
sequence = pbx.assign(dihedrals)
all_comments.append(comment)
all_sequences.append(sequence)
except FloatingPointError:
log.error("The computation of angles produced NaN. "
"This typically means there are issues with "
"some residues coordinates. "
f"Check your input file ({comment})")
log.info(f"There are {len(all_sequences)} sequences of length "
f"{len(all_sequences[0])}")
log.info("Calculating the Mutual Information matrix ...")
MI_matrix = mutual_information_matrix(all_sequences)
# Write to a file
log.info(f"Writing the matrix as {options.output} ...")
df = pd.DataFrame(MI_matrix)
df.to_csv(options.output)
# Add option in CLI, centrality
# Creating a network.
if options.network:
log.info("Creating a network ...")
PB_graph = interaction_graph(MI_matrix)
log.info(f"Writing the network as {options.network} ...")
# Write the graph to GML format.
nx.write_gml(PB_graph, path=options.network)