1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
|
import os
import h5py
import numpy as np
import pytest
import cooler
from cooler import _balance
testdir = os.path.dirname(os.path.realpath(__file__))
@pytest.mark.parametrize(
"fp,tol",
[(os.path.join(testdir, "data", "hg19.GM12878-MboI.matrix.2000kb.cool"), 1e-2)],
)
def test_balancing_genomewide(fp, tol):
clr = cooler.Cooler(fp)
weights, stats = _balance.iterative_correction(
clr, ignore_diags=1, min_nnz=10, tol=tol
)
# Extract matrix and apply weights
mat = cooler.Cooler(fp).matrix(balance=False, sparse=True)[:, :]
mat.data = weights[mat.row] * weights[mat.col] * mat.data
arr = mat.toarray()
# Re-apply bin level filters
mask = np.isnan(weights)
arr[mask, :] = 0
arr[:, mask] = 0
# Apply diagonal filter
np.fill_diagonal(arr, 0)
# Check that the balanced marginal is flat
marg = np.sum(arr, axis=0)
var = np.var(marg[marg != 0])
assert var < tol
# Check that the balanced marginal is unity
conv_marg = marg[~np.isnan(marg)].mean()
err_marg = marg[~np.isnan(marg)].std()
assert np.isclose(conv_marg, 1, atol=err_marg)
@pytest.mark.filterwarnings("ignore")
@pytest.mark.parametrize(
"fp,tol",
[(os.path.join(testdir, "data", "hg19.GM12878-MboI.matrix.2000kb.cool"), 1e-2)],
)
def test_balancing_cisonly(fp, tol):
with h5py.File(fp, "r") as h5:
clr = cooler.Cooler(h5)
chrom_offsets = h5["indexes/chrom_offset"][:]
weights, stats = _balance.iterative_correction(
clr, ignore_diags=1, min_nnz=10, tol=tol, cis_only=True
)
# Extract matrix and apply weights
mat = cooler.Cooler(fp).matrix(balance=False, sparse=True)[:, :]
mat.data = weights[mat.row] * weights[mat.col] * mat.data
arr = mat.toarray()
# Re-apply bin level filters
mask = np.isnan(weights)
arr[mask, :] = 0
arr[:, mask] = 0
# Apply diagonal filter
np.fill_diagonal(arr, 0)
# Filter out trans data
spans = list(zip(chrom_offsets[:-1], chrom_offsets[1:]))
from scipy.linalg import block_diag
blocks = [np.ones((hi - lo,) * 2) for lo, hi in spans]
mask = block_diag(*blocks).astype(bool)
arr[~mask] = 0
# Check that the balanced marginal is flat
marg = np.sum(arr, axis=0)
for lo, hi in spans:
m = marg[lo:hi]
m = m[m != 0]
if len(m):
print(lo, hi)
var = np.var(m[m != 0])
assert var < tol
# Check that the balanced marginal is unity
conv_marg = m[~np.isnan(m)].mean()
err_marg = m[~np.isnan(m)].std()
assert np.isclose(conv_marg, 1, atol=err_marg)
@pytest.mark.filterwarnings("ignore")
@pytest.mark.parametrize(
"fp,tol",
[(os.path.join(testdir, "data", "hg19.GM12878-MboI.matrix.2000kb.cool"), 1e-2)],
)
def test_balancing_transonly(fp, tol):
with h5py.File(fp, "r") as h5:
clr = cooler.Cooler(h5)
chrom_offsets = h5["indexes/chrom_offset"][:]
weights, stats = _balance.iterative_correction(
clr, ignore_diags=1, min_nnz=10, tol=tol, trans_only=True
)
# Extract matrix and apply weights
mat = cooler.Cooler(fp).matrix(balance=False, sparse=True)[:, :]
mat.data = weights[mat.row] * weights[mat.col] * mat.data
arr = mat.toarray()
# Re-apply bin level filters
mask = np.isnan(weights)
arr[mask, :] = 0
arr[:, mask] = 0
# Filter out cis data
for lo, hi in zip(chrom_offsets[:-1], chrom_offsets[1:]):
arr[lo:hi, lo:hi] = np.nan
# Check that the balanced marginal is flat
marg = np.nansum(arr, axis=0)
var = np.nanvar(marg[marg != 0])
assert var < tol
# Check that the balanced marginal is unity
conv_marg = marg[~np.isnan(marg)].mean()
err_marg = marg[~np.isnan(marg)].std()
assert np.isclose(conv_marg, 1, atol=err_marg)
@pytest.mark.parametrize(
"fp,tol",
[(os.path.join(testdir, "data", "toy.symm.upper.2.cool"), 1e-2)],
)
def test_balancing_other_options(fp, tol):
clr = cooler.Cooler(fp)
weights, stats = _balance.iterative_correction(
clr, ignore_diags=1, min_nnz=10, tol=tol, x0=np.random.rand(len(clr.bins()))
)
weights, stats = _balance.iterative_correction(
clr,
chunksize=3,
ignore_diags=1,
min_nnz=10,
tol=tol,
)
weights, stats = _balance.iterative_correction(
clr,
ignore_diags=1,
min_nnz=10,
min_count=1,
mad_max=12,
tol=tol,
)
weights, stats = _balance.iterative_correction(
clr, ignore_diags=1, min_nnz=10, tol=tol, blacklist=[0, 4]
)
|
