def ConsurfParser(input, input_format=None):
contents = input.split('\n')
output = []
for line in contents:
line = line.lstrip().split()
if len(line) < 4 or not line[0].isnumeric(
) or not line[3][0].isnumeric():
continue
else:
score = int(line[3][0])
if score > 9:
raise InvalidFormat(
'Unable to parse prediction on consurf file: score above 9 detected!'
)
else:
output.append(score)
if not output:
raise InvalidFormat('Unable to parse prediction on consurf file')
else:
return output
def CCMpredParser(input, input_format=None):
contents = input.split('\n')
output = []
for res_1, line in enumerate(contents, 1):
line = line.lstrip().split()
if not line or line[0].isalpha() or len(line) == 1 or '#' in line[0]:
continue
for res_2, raw_score in enumerate(line, 1):
try:
raw_score = float(raw_score)
except ValueError:
raise InvalidFormat('Unable to parse contacts')
if raw_score == '' or raw_score < 0.1:
continue
seq_distance = abs(res_1 - res_2)
if seq_distance >= 5:
contact = [res_1, res_2, raw_score]
contact[:2] = sorted(contact[:2], reverse=True)
output.append((tuple(contact[:2]), contact[2]))
if not output:
raise InvalidFormat('Unable to parse contacts')
else:
unique_contacts = get_unique_contacts(output)
return unique_contacts
def CASPRR2Parser(input, input_format=None):
contents = input.split('\n')
output = []
res_1_idx = 0
res_2_idx = 1
raw_score_idx = 2
line_size = 13
for idx, line in enumerate(contents):
line = line.lstrip().rstrip().split()
if not line or len(line) != line_size or not line[res_1_idx].isdigit() or not line[res_2_idx].isdigit():
continue
res_1 = int(line[res_1_idx])
res_2 = int(line[res_2_idx])
seq_distance = res_1 - res_2
if abs(seq_distance) >= 5:
raw_score = float(line[raw_score_idx])
distance_probabilities = [float(p) for p in line[raw_score_idx + 1:]]
distance_score = max(distance_probabilities)
distance_bin = distance_probabilities.index(distance_score)
contact = [res_1, res_2, raw_score, distance_bin, distance_score]
contact[:2] = sorted(contact[:2], reverse=True)
output.append((tuple(contact[:2]), *contact[2:]))
if not output:
raise InvalidFormat('Unable to parse CASPRR_MODE_2 file')
else:
unique_contacts = get_unique_distances(output)
if any([p for p in unique_contacts[1:] if p[3] > 9 or p[4] > 1]):
raise InvalidFormat('Unable to parse CASPRR_MODE_2 file')
return unique_contacts
def PDBParser(input, input_format=None):
try:
parser = BioPDBParser().get_structure('pdb', io.StringIO(input))
chain = list(parser.get_chains())[0]
remove_atoms(chain)
contacts = get_chain_contacts(chain)
except:
raise InvalidFormat('Unable to parse contacts')
if not contacts:
raise InvalidFormat('Unable to parse contacts')
output = ["PDB"]
output += sorted(contacts, key=itemgetter(2), reverse=True)
return output
def ContactParser(input, input_format):
contents = input.split('\n')
output = []
res_1_idx = FieldResidueOneContactFormats.__getattr__(input_format).value
res_2_idx = FieldResidueTwoContactFormats.__getattr__(input_format).value
raw_score_idx = FieldRawScoreContactFormats.__getattr__(input_format).value
line_size = LineSizeContactFormats.__getattr__(input_format).value
regex = FieldSeparatorContactFormats.__getattr__(input_format).value
for idx, line in enumerate(contents):
line = line.lstrip().rstrip()
line = re.split(regex, line)
if not line or len(line) < line_size or not line[res_1_idx].isdigit() or not line[res_2_idx].isdigit():
continue
res_1 = int(line[res_1_idx])
res_2 = int(line[res_2_idx])
seq_distance = res_1 - res_2
if abs(seq_distance) >= 5:
if raw_score_idx is not None:
raw_score = float(line[raw_score_idx])
else:
raw_score = 0
contact = [res_1, res_2, raw_score]
contact[:2] = sorted(contact[:2], reverse=True)
output.append((tuple(contact[:2]), contact[2]))
if not output:
raise InvalidFormat('Unable to parse contacts')
else:
unique_contacts = get_unique_contacts(output)
return unique_contacts
def guess_psipred_format(contents):
for line in contents:
if '# PSIPRED VFORMAT' in line:
return Ss2Parser
elif '# PSIPRED HFORMAT' in line:
return HorizParser
raise InvalidFormat('Unable to guess psipred file format')
def Ss2Parser(contents):
output = []
for line in contents:
line = line.split()
if len(line) != 6 or line[0] == '#':
continue
elif line[2] == 'H':
output.append(SecondaryStructureStates.HELIX.value)
elif line[2] == 'C':
output.append(SecondaryStructureStates.COIL.value)
elif line[2] == 'E':
output.append(SecondaryStructureStates.SHEET.value)
else:
raise InvalidFormat(
'Invalid secondary structure element {}'.format(line[2]))
if not output:
raise InvalidFormat('Unable to parse prediction in psipred file')
else:
return output
def A3mParser(input, input_format=None):
sequences, seq_lenght, n_sequences = extract_sequences(input)
if sequences is None or len(
sequences) <= 1 or n_sequences < 0.1 or seq_lenght < 1:
raise InvalidFormat('Unable to parse contents of A3M MSA file')
residue_count = [0 for x in range(1, seq_lenght + 1)]
try:
for sequence in sequences:
idx = 0
for residue in sequence:
if residue.islower():
continue
elif residue != '-':
residue_count[idx] += 1
idx += 1
except IndexError:
raise InvalidFormat('Unable to parse the A3M MSA file')
norm = [int(round(x / n_sequences)) for x in residue_count]
return norm
def HorizParser(contents):
output = []
for line in contents:
if 'Pred: ' in line and line[0] != '#':
prediction = line.split()[-1]
for residue in prediction:
if residue == 'H':
output.append(SecondaryStructureStates.HELIX.value)
elif residue == 'C':
output.append(SecondaryStructureStates.COIL.value)
elif residue == 'E':
output.append(SecondaryStructureStates.SHEET.value)
else:
raise InvalidFormat(
'Invalid secondary structure element {}'.format(
residue))
if not output:
raise InvalidFormat('Unable to parse prediction in psipred file')
else:
return output
def IupredParser(input, input_format=None):
contents = input.split('\n')
output = []
for line in contents:
line = line.lstrip().split()
if len(line) < 1 or line[0] == '#' or len(line) < 3:
continue
else:
try:
score = float(line[2])
except ValueError:
raise InvalidFormat('Invalid score field {}'.format(line[2]))
if score >= 0.5:
output.append(DisorderStates.DISORDER.value)
else:
output.append(DisorderStates.ORDER.value)
if not output:
raise InvalidFormat('Unable to parse prediction on iupred file')
else:
return output
def NpzParser(input, input_format=None):
output = []
content_type, content_string = input.split(',')
try:
decoded = base64.b64decode(content_string)
archive = np.load(io.BytesIO(decoded), allow_pickle=True)
array = archive['dist']
tmp_output = parse_array(array)
except (OSError, KeyError, IndexError) as e:
raise InvalidFormat('Unable to parse distance NPZ file')
for contact in tmp_output:
# contact = [res_1, res_2, raw_score, distance_bin, distance_score]
contact[:2] = sorted(contact[:2], reverse=True)
output.append((tuple(contact[:2]), *contact[2:]))
if not output:
raise InvalidFormat('Unable to parse NPZ file')
else:
unique_contacts = get_unique_distances(output)
if any([p for p in unique_contacts[1:] if p[3] > 9 or p[4] > 1]):
raise InvalidFormat('Unable to parse trRosetta_NPZ file')
return unique_contacts
def TopconsParser(input, input_format=None):
contents = input.split('\n')
try:
topcons_prediction = contents[contents.index('TOPCONS predicted topology:') + 1].rstrip()
except ValueError as e:
raise InvalidFormat
output = []
for residue in topcons_prediction.rstrip().lstrip():
if residue == 'i':
output.append(MembraneStates.INSIDE.value)
elif residue == 'o':
output.append(MembraneStates.OUTSIDE.value)
elif residue == 'M':
output.append(MembraneStates.INSERTED.value)
else:
raise InvalidFormat('Invalid residue topology {}'.format(residue))
if not output:
raise InvalidFormat('Unable to parse prediction in topcons file')
else:
return output
def MappredParser(input, input_format=None):
contents = input.split('\n')
output = []
res_1_idx = 0
res_2_idx = 1
line_size = 36
for idx, line in enumerate(contents):
line = line.lstrip().rstrip().split()
if not line or len(line) < line_size or not line[res_1_idx].isdigit() or not line[res_2_idx].isdigit():
continue
res_1 = int(line[res_1_idx])
res_2 = int(line[res_2_idx])
seq_distance = res_1 - res_2
if abs(seq_distance) >= 5:
line = [float(prob) for prob in line[2:]]
raw_score = sum([prob for prob in line[:9]])
distance_probabilities = [line[0]]
distance_probabilities += [sum(line[x:x+4]) for x in range(1, 30, 4)]
distance_probabilities.append(line[-1])
distance_score = max(distance_probabilities)
distance_bin = distance_probabilities.index(distance_score)
contact = [res_1, res_2, raw_score, distance_bin, distance_score]
contact[:2] = sorted(contact[:2], reverse=True)
output.append((tuple(contact[:2]), *contact[2:]))
if not output:
raise InvalidFormat('Unable to parse MapPred file')
else:
unique_contacts = get_unique_distances(output)
if any([p for p in unique_contacts[1:] if p[3] > 9 or p[4] > 1]):
raise InvalidFormat('Unable to parse MapPred file')
return unique_contacts
def ComsatParser(input):
contents = input.split('\n')
output = []
for line in contents:
line = line.lstrip().split()
if not line or line[0].isalpha():
continue
elif line[0].isdigit() and line[2].isdigit() and len(line) >= 5:
if abs(int(line[0]) - int(line[2])) >= 5:
output.append((int(line[0]), int(line[2]), (0)))
if not output:
raise InvalidFormat('Unable to parse contacts')
else:
output = sorted(output, key=itemgetter(2), reverse=True)
return output
def CCMpredParser(input):
contents = input.split('\n')
output = []
data = []
for line in contents:
line = line.lstrip().split()
if not line or line[0].isalpha() or len(line) == 1:
continue
else:
data.append(line)
for res_1, score_array in enumerate(data, 1):
for res_2, score in enumerate(score_array, 1):
if abs((res_1) - int(res_2)) >= 5 and score != '' and float(score) > 0:
output.append((int(res_1), int(res_2), float(score)))
if not output:
raise InvalidFormat('Unable to parse contacts')
else:
output = sorted(output, key=itemgetter(2), reverse=True)
return output