def process_master_data_no_gap(self, masfile: Path, name1: str,
name2: str) -> pd.DataFrame:
"""Get length data from the MASTER matches.
"""
def cutter(row):
match = row['match']
# MASTER starts match count at 0!
return row['abego'][match[0][0]:match[0][-1] + 1], '-', 0
if masfile.with_suffix('.csv').is_file():
df = pd.read_csv(masfile.with_suffix('.csv'))
df['match'] = df['match'].apply(literal_eval)
return df
dfloop = parse_master_file(masfile)
dfloop = dfloop.merge(self.abegos,
on=['pdb',
'chain']).merge(self.fragments,
on=['pdb',
'chain']).dropna()
dfloop[['abego', 'loop',
'loop_length']] = dfloop.apply(cutter,
axis=1,
result_type='expand')
dfloop = dfloop.iloc[:self.top_loops]
dfloop['length_count'] = dfloop.loop_length.map(
dfloop.loop_length.value_counts())
dfloop.drop(columns=['pds_path']).to_csv(
masfile.with_suffix('.all.csv'), index=False)
finaldf = dfloop.sort_values('rmsd').drop_duplicates(['loop'])
df = finaldf.drop(columns=['pds_path'])
df.to_csv(masfile.with_suffix('.csv'), index=False)
return df
def test_master(self):
df = parse_master_file(os.path.join(self.dirpath, 'master.search'),
max_rmsd=1.4,
piece_count=2,
shift_0=True)
assert df.rmsd.max() == 1.3967
assert df.shape == (42, 5)
assert df.iloc[-1].match == [[34, 40], [42, 48]]
def process_master_data(self, masfile: Path, name1: str, name2: str,
hairpin: bool) -> pd.DataFrame:
"""Get length data from the MASTER matches.
"""
def cutter(row):
match = row['match']
# MASTER starts match count at 0!
loop = row['abego'][match[0][1] + 1:match[1][0]]
return row['abego'][match[0][0]:match[1][1] + 1], loop, len(loop)
if masfile.with_suffix('.csv').is_file():
df = pd.read_csv(masfile.with_suffix('.csv'))
df['match'] = df['match'].apply(literal_eval)
return df
dfloop = parse_master_file(masfile)
dfloop = dfloop.merge(self.abegos,
on=['pdb',
'chain']).merge(self.fragments,
on=['pdb',
'chain']).dropna()
dfloop[['abego', 'loop',
'loop_length']] = dfloop.apply(cutter,
axis=1,
result_type='expand')
dfloop = dfloop.iloc[:self.top_loops]
dfloop['length_count'] = dfloop.loop_length.map(
dfloop.loop_length.value_counts())
dfloop.drop(columns=['pds_path']).to_csv(
masfile.with_suffix('.all.csv'), index=False)
finaldf = dfloop.sort_values('rmsd').drop_duplicates(['loop'])
pick = 0
if hairpin and 2 in finaldf['loop_length'].values:
pick = 2
else:
pick = finaldf[finaldf['length_count'] ==
finaldf['length_count'].max()]['loop_length'].min()
finaldf = finaldf[finaldf['loop_length'] == pick]
TBPlot.plot_loop_length_distribution(self.log, dfloop, pick,
masfile.with_suffix(''),
f'loop {name1} <-> {name2}')
df = finaldf.drop(columns=['pds_path'])
df.to_csv(masfile.with_suffix('.csv'), index=False)
return df
def main(options):
"""
"""
# Load MASTER search data.
masterdf = parse_master_file(options.master, shift_0=True)
# Case data
case = Case(Path(options.case))
# Get connectivities
sse = case.connectivities_str[0].split('.')
# Get flips
flip = cycle([
case['configuration.flip_first'], not case['configuration.flip_first']
])
flip = [next(flip) for _ in range(len(sse))]
# Select only the present ones.
present = [sse.index(i) for i in options.present.split('.')]
sse = list(itemgetter(*present)(sse))
flip = list(itemgetter(*present)(flip))
# Geometric properties retrieval
masterdf = process_master_geometries(masterdf, sse, flip)
# Output data
masterdf.to_csv(str(options.out) + '.csv', index=False)
def process_master_data(self, masfile: Path, names: List,
loop_lengths: str,
loop_orders: str) -> pd.DataFrame:
"""Get length data from the MASTER matches.
"""
def cutter(row, num):
match = row['match']
# MASTER starts match count at 0!
loop = row['abego'][match[num][1] + 1:match[num + 1][0]]
return row['abego'][match[num][0]:match[num + 1][1] +
1], loop, len(loop), int(match[num][0]), int(
match[num + 1][1])
if masfile.with_suffix('.csv').is_file():
df = pd.read_csv(masfile.with_suffix('.csv'))
df['match'] = df['match'].apply(literal_eval)
return df
llens = loop_lengths.split(';')
pnames = [(names[i], names[i + 1]) for i in range(len(names) - 1)]
lorder = loop_orders.split(';')
dfloop = parse_master_file(masfile)
dfloop = dfloop.merge(self.abegos,
on=['pdb',
'chain']).merge(self.fragments,
on=['pdb',
'chain']).dropna()
container = []
for k, (pname, llen, lord) in enumerate(zip(pnames, llens, lorder)):
if lord is 'x': # skip regions that are not of interest
continue
# set up
nfolder = masfile.parent.absolute().joinpath(
f'loop{int(lord):02d}')
nfolder.mkdir(parents=True, exist_ok=True)
masfile2 = str(nfolder.joinpath(f'jump{int(lord):02d}'))
dfloop_copy = dfloop.copy()
self.log.info(
f'Current jump is {pname[0], pname[1]} of order {lord} located at {k}'
)
dfloop_copy[['abego', 'loop', 'loop_length', 'start',
'stop']] = dfloop_copy.apply(cutter,
num=k,
axis=1,
result_type='expand')
dfloop_copy = dfloop_copy.iloc[:self.top_loops]
dfloop_copy['length_count'] = dfloop_copy.loop_length.map(
dfloop_copy.loop_length.value_counts())
finaldf = dfloop_copy.sort_values('rmsd').drop_duplicates(['loop'])
is_hairpin = self.check_hairpin(pname[0], pname[1])
if self.pick_by == 'minimal':
pick = finaldf['loop_length'].min()
if self.hairpins_2 == True and is_hairpin == True:
pick = 2
finaldf = finaldf[finaldf['loop_length'] == pick]
else:
pick = finaldf[
finaldf['length_count'] ==
finaldf['length_count'].max()]['loop_length'].min()
if self.hairpins_2 == True and is_hairpin == True:
pick = 2
finaldf = finaldf[finaldf['loop_length'] == pick]
TBPlot.plot_loop_length_distribution(
self.log, dfloop_copy, pick, Path(masfile2),
f'loop {pname[0]} <-> {pname[1]}')
df = finaldf.drop(columns=['pds_path'])
df = df.assign(order=[int(lord)] * len(df))
df.to_csv(masfile2 + '.csv', index=False)
container.append(df)
df = pd.concat(container).sort_values('order')
df.to_csv(masfile.with_suffix('.all.csv'), index=False)
return df