def rename_weights(cooleruri, name_map):
cool_path, group_path = parse_cooler_uri(cooleruri)
with h5py.File(cool_path, 'r+') as h5:
grp = h5[group_path]
h5opts = dict(compression='gzip', compression_opts=6)
for old_name, new_name in name_map.items():
# add the bias column to the file
weights = grp['bins'][old_name][()].copy()
grp['bins'].create_dataset(new_name, data=weights, **h5opts)
del grp['bins'][old_name]
def test_parse_cooler_uri():
for uri in [
'/foo/bar/baz.mcool::resolutions/1000',
'/foo/bar/baz.mcool::/resolutions/1000'
]:
fp, gp = util.parse_cooler_uri(uri)
assert fp == '/foo/bar/baz.mcool'
assert gp == '/resolutions/1000'
for uri in ['/foo/bar/baz.cool', '/foo/bar/baz.cool::/']:
fp, gp = util.parse_cooler_uri(uri)
assert fp == '/foo/bar/baz.cool'
assert gp == '/'
for uri in [
'/foo/bar/baz.cool::/a/b::c.cool',
]:
with pytest.raises(ValueError):
util.parse_cooler_uri(uri)
def matrix_balance(cool_uri,
nproc=1,
chunksize=int(1e7),
mad_max=5,
min_nnz=10,
min_count=0,
ignore_diags=1,
tol=1e-5,
max_iters=1000):
'''
Perform separate matrix balancing for regions with different copy numbers
and output the bias vector in the "sweight" column.
'''
cool_path, group_path = util.parse_cooler_uri(cool_uri)
# Overwrite the existing sweight column
with h5py.File(cool_path, 'r+') as h5:
grp = h5[group_path]
if 'sweight' in grp['bins']:
del grp['bins']['sweight']
clr = Cooler(cool_uri)
try:
if nproc > 1:
pool = balance.Pool(nproc)
map_ = pool.imap_unordered
else:
map_ = map
bias, stats = iterative_correction(clr,
chunksize=chunksize,
tol=tol,
min_nnz=min_nnz,
min_count=min_count,
mad_max=mad_max,
max_iters=max_iters,
ignore_diags=ignore_diags,
rescale_marginals=True,
use_lock=False,
map=map_)
finally:
if nproc > 1:
pool.close()
if not stats['converged']:
logger.error('Iteration limit reached without convergence')
logger.error('Storing final result. Check log to assess convergence.')
with h5py.File(cool_path, 'r+') as h5:
grp = h5[group_path]
# add the bias column to the file
h5opts = dict(compression='gzip', compression_opts=6)
grp['bins'].create_dataset('sweight', data=bias, **h5opts)
grp['bins']['sweight'].attrs.update(stats)
def store_weights(cooleruri, bias, weightname):
'''
stores an iterable of values as a new weight column in the given cooleruri
with name set to wightname. code taken from cooler's cooler balance see also
https://github.com/mirnylab/cooler/blob/master/cooler/cli/balance.py
:param cooleruri: uri to a given cooler
:param bias: iterable containing balancing weights for each genomic bin
:param weightname: name of the weight column
:return: None
'''
cool_path, group_path = parse_cooler_uri(cooleruri)
with h5py.File(cool_path, 'r+') as h5:
grp = h5[group_path]
# add the bias column to the file
h5opts = dict(compression='gzip', compression_opts=6)
grp['bins'].create_dataset(weightname, data=bias, **h5opts)
def check_weight(cooleruri, weight_name):
'''
checks if weight_name already exist in cooler file
:param cooleruri: uri to a given cooleruri
:param weight_name: name of the weight to check for
:return: True if weight already in cooler else False
'''
cool_path, group_path = parse_cooler_uri(cooleruri)
weight_exists = False
with h5py.File(cool_path, 'r+') as h5:
grp = h5[group_path]
if grp['bins'].get(weight_name):
weight_exists = True
return weight_exists
def assign_cnv(self, cooler_uri):
cooler_lib = cooler.Cooler(cooler_uri)
ref_pre = find_chrom_pre(cooler_lib.chromnames)
cnv = np.r_[[]]
for ref_k in cooler_lib.chromnames: # line with bin table
bias = cooler_lib.bins().fetch(ref_k)['weight'].values
c = ref_k.lstrip(ref_pre)
if not c in self.bin_cnv:
pre = np.zeros(len(bias))
cnv = np.r_[cnv, pre]
continue
pre = self.bin_cnv[c]
if len(bias) <= pre.size:
pre = pre[:len(bias)]
else:
add = np.zeros(len(bias) - pre.size)
pre = np.r_[pre, add]
mask = np.isnan(bias) | (bias == 0)
pre[mask] = 0
cnv = np.r_[cnv, pre]
cnvi = rankdata(cnv,
method='dense') - 1 # indices for quick Bias retrival
# pre-check the CNV column
cool_path, group_path = util.parse_cooler_uri(cooler_uri)
with h5py.File(cool_path, 'r+') as h5:
grp = h5[group_path]
if 'CNV' in grp['bins']:
del grp['bins']['CNV'] # Overwrite the CNV column
del grp['bins']['CNVI']
with h5py.File(cool_path, 'r+') as h5:
grp = h5[group_path]
h5opts = dict(compression='gzip', compression_opts=6)
grp['bins'].create_dataset('CNV', data=cnv, **h5opts)
grp['bins'].create_dataset('CNVI', data=cnvi, dtype='i1', **h5opts)
del cnv, cnvi
def save_weight(coolfile, chroms, weights, bal_info={}, col_name='weight'):
"""
coolfile - File to add weights too
chroms - List of chromosomes to modify
weights - List of weight vectors to write to corresponding chromosomes
bal_info - Dict to add as attributes of the weight column"""
cool_path, group_path = util.parse_cooler_uri(coolfile)
c = cooler.Cooler(coolfile)
with h5py.File(cool_path, 'r+') as h5:
grp = h5[group_path]
# add the bias column to the file
h5opts = dict(compression='gzip', compression_opts=6)
allweights = np.asarray([np.nan]*grp['bins']['chrom'].shape[0])
if col_name not in grp['bins'].keys():
grp['bins'].create_dataset(col_name, data=allweights, **h5opts)
for chrom, weight in zip(chroms, weights):
s, e = c.extent(chrom)
grp['bins'][col_name][s:e] = weight
grp['bins'][col_name].attrs.update(bal_info)
# grp['bins']['weight'][grp['bins']['chrom']==chrom]=weights
return
def balance(cool_uri, nproc=1, chunksize=int(1e7), mad_max=5, min_nnz=10,
min_count=0, ignore_diags=1, tol=1e-5, max_iters=200):
"""
Cooler contact matrix balancing.
Parameters
----------
cool_uri : str
URI of cooler group.
nproc : int
Number of processes. (Default: 1)
"""
cool_path, group_path = parse_cooler_uri(cool_uri)
# pre-check the weight column
with h5py.File(cool_path, 'r+') as h5:
grp = h5[group_path]
if 'weight' in grp['bins']:
del grp['bins']['weight'] # Overwrite the weight column
log.info('Balancing {0}'.format(cool_uri))
clr = Cooler(cool_uri)
try:
if nproc > 1:
pool = Pool(nproc)
map_ = pool.imap_unordered
else:
map_ = map
if clr.info['metadata']['onlyIntra']=='True':
onlyIntra = True
else:
onlyIntra = False
bias, stats = ice.iterative_correction(
clr,
chunksize=chunksize,
cis_only=onlyIntra,
trans_only=False,
tol=tol,
min_nnz=min_nnz,
min_count=min_count,
blacklist=None,
mad_max=mad_max,
max_iters=max_iters,
ignore_diags=ignore_diags,
rescale_marginals=True,
use_lock=False,
map=map_)
finally:
if nproc > 1:
pool.close()
if not stats['converged']:
log.error('Iteration limit reached without convergence')
log.error('Storing final result. Check log to assess convergence.')
with h5py.File(cool_path, 'r+') as h5:
grp = h5[group_path]
# add the bias column to the file
h5opts = dict(compression='gzip', compression_opts=6)
grp['bins'].create_dataset('weight', data=bias, **h5opts)
grp['bins']['weight'].attrs.update(stats)
def balance(cool_uri, nproc=1, chunksize=int(1e7), mad_max=5, min_nnz=10,
min_count=0, ignore_diags=1, tol=1e-5, max_iters=200):
"""
Cooler contact matrix balancing.
Parameters
----------
cool_uri : str
URI of cooler group.
nproc : int
Number of processes. (Default: 1)
"""
cool_path, group_path = parse_cooler_uri(cool_uri)
# pre-check the weight column
with h5py.File(cool_path, 'r+') as h5:
grp = h5[group_path]
if 'weight' in grp['bins']:
del grp['bins']['weight'] # Overwrite the weight column
log.info('Balancing {0}'.format(cool_uri))
clr = Cooler(cool_uri)
try:
if nproc > 1:
pool = Pool(nproc)
map_ = pool.imap_unordered
else:
map_ = map
if clr.info['metadata']['onlyIntra']=='True':
onlyIntra = True
else:
onlyIntra = False
bias, stats = ice.iterative_correction(
clr,
chunksize=chunksize,
cis_only=onlyIntra,
trans_only=False,
tol=tol,
min_nnz=min_nnz,
min_count=min_count,
blacklist=None,
mad_max=mad_max,
max_iters=max_iters,
ignore_diags=ignore_diags,
rescale_marginals=True,
use_lock=False,
map=map_)
finally:
if nproc > 1:
pool.close()
if not stats['converged']:
log.error('Iteration limit reached without convergence')
log.error('Storing final result. Check log to assess convergence.')
with h5py.File(cool_path, 'r+') as h5:
grp = h5[group_path]
# add the bias column to the file
h5opts = dict(compression='gzip', compression_opts=6)
grp['bins'].create_dataset('weight', data=bias, **h5opts)
grp['bins']['weight'].attrs.update(stats)