def test_by_annotator(self, new_pipegraph_no_qc):
genome = get_human_22_fake_genome()
start = 17750239
df = pd.DataFrame(
[
{
"chr": "chr22",
"start": start,
"stop": start + 1000,
},
{
"chr": "chr22",
"start": start + 20000,
"stop": start + 20000 + 1000,
},
{
"chr": "chr22",
"start": start + 30000,
"stop": start + 30000 + 1000,
},
]
)
lane1 = mbf_align.lanes.AlignedSample(
"one",
mbf_sampledata.get_sample_path("mbf_align/chipseq_chr22.bam"),
genome,
False,
None,
)
lanes = {lane1.name: lane1}
raw_data = {
lane1.name: np.array(
[
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
]
)
}
plot_regions = mbf_genomics.regions.GenomicRegions(
"testregions", lambda: df, [], genome
)
class FakeAnno(mbf_genomics.annotator.Annotator):
columns = ["colA"]
def calc(self, df):
return pd.Series([1, 3, 2])
o = order.ByAnnotator(FakeAnno())
ppg.JobGeneratingJob("shu", lambda: None).depends_on(
o.get_dependencies(plot_regions, lanes)[0]
)
ppg.run_pipegraph()
plot_regions._load()
norm_data = norm.AsIs().calc(lanes, raw_data)
res_order, clusters = o.calc(plot_regions, lanes, raw_data, norm_data)
assert (res_order == [0, 2, 1]).all()
def test_by_column(self, new_pipegraph_no_qc):
genome = get_human_22_fake_genome()
start = 17750239
df = pd.DataFrame(
[
{
"chr": "chr22",
"start": start,
"stop": start + 1000,
"colA": "a",
},
{
"chr": "chr22",
"start": start + 20000,
"stop": start + 20000 + 1000,
"colA": "c",
},
{
"chr": "chr22",
"start": start + 30000,
"stop": start + 30000 + 1000,
"colA": "b",
},
]
)
lane1 = mbf_align.lanes.AlignedSample(
"one",
mbf_sampledata.get_sample_path("mbf_align/chipseq_chr22.bam"),
genome,
False,
None,
)
lanes = {lane1.name: lane1}
o = order.ByAnnotator("colA", func=lambda x: [ord(y) for y in x])
raw_data = {
lane1.name: np.array(
[
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
]
)
}
plot_regions = mbf_genomics.regions.GenomicRegions(
"testregions", lambda: df, [], genome
)
ppg.JobGeneratingJob("shu", lambda: None).depends_on(plot_regions.load())
ppg.run_pipegraph()
plot_regions._load()
norm_data = norm.AsIs().calc(lanes, raw_data)
res_order, clusters = o.calc(plot_regions, lanes, raw_data, norm_data)
assert (res_order == [0, 2, 1]).all()
def test_smooth(self, new_pipegraph_no_qc):
genome = get_human_22_fake_genome()
df = pd.DataFrame(
[
{
"chr": "chr22",
"start": 36925 * 1000 - 1000,
"stop": 36925 * 1000 + 1000,
},
{
"chr": "chr22",
"start": 31485 * 1000 - 2000,
"stop": 31485 * 1000 + 2000,
},
{"chr": "chr22", "start": 41842 * 1000, "stop": (41842 * 1000) + 1},
]
)
plot_regions = mbf_genomics.regions.GenomicRegions(
"testregions", lambda: df, [], genome
)
lane1 = mbf_align.lanes.AlignedSample(
"one",
mbf_sampledata.get_sample_path("mbf_align/chipseq_chr22.bam"),
genome,
False,
None,
)
lane2 = mbf_align.lanes.AlignedSample(
"two",
mbf_sampledata.get_sample_path("mbf_align/chipseq_chr22.bam"),
genome,
False,
None,
)
h = mbf_heatmap.chipseq.Heatmap(
plot_regions,
[lane1, lane2],
region_strategy=regions.RegionFromCenter(1000),
smoothing_strategy=smooth.SmoothExtendedReads(),
)
fn = "test.png"
h.plot(fn, norm.AsIs(), order.FirstLaneSum())
ppg.run_pipegraph()
assert_image_equal(fn)
def test_simple(self, new_pipegraph_no_qc):
genome = get_human_22_fake_genome()
start = 17750239
df = pd.DataFrame(
[
{"chr": "chr22", "start": start, "stop": start + 1000},
{"chr": "chr22", "start": start + 20000, "stop": start + 20000 + 1000},
{"chr": "chr22", "start": start + 30000, "stop": start + 30000 + 1000},
]
)
plot_regions = mbf_genomics.regions.GenomicRegions(
"testregions", lambda: df, [], genome
)
lane1 = mbf_align.lanes.AlignedSample(
"one",
mbf_sampledata.get_sample_path("mbf_align/chipseq_chr22.bam"),
genome,
False,
None,
)
lane2 = mbf_align.lanes.AlignedSample(
"two",
mbf_sampledata.get_sample_path("mbf_align/chipseq_chr22.bam"),
genome,
False,
None,
)
h = mbf_heatmap.chipseq.Heatmap(
plot_regions,
[lane1, lane2],
region_strategy=regions.RegionAsIs(),
smoothing_strategy=smooth.SmoothRaw(),
)
fn = "test.png"
h.plot(fn, norm.AsIs(), order.AsIs())
ppg.run_pipegraph()
assert_image_equal(fn)
def test_ithlane_max(self, new_pipegraph):
genome = get_human_22_fake_genome()
start = 17750239
df = pd.DataFrame(
[
{"chr": "chr22", "start": start, "stop": start + 1000},
{"chr": "chr22", "start": start + 20000, "stop": start + 20000 + 1000},
{"chr": "chr22", "start": start + 30000, "stop": start + 30000 + 1000},
]
)
lane1 = mbf_align.lanes.AlignedSample(
"one",
mbf_sampledata.get_sample_path("mbf_align/chipseq_chr22.bam"),
genome,
False,
None,
)
lane2 = mbf_align.lanes.AlignedSample(
"two",
mbf_sampledata.get_sample_path("mbf_align/chipseq_chr22.bam"),
genome,
False,
None,
)
with pytest.raises(AttributeError):
order.IthLaneMax(lane1.name)
o = order.IthLaneMax(1)
# raw_data = {lane1.name: smooth.SmoothRaw().calc(df, lane1)}
raw_data = {
lane1.name: np.array(
[
[0, 0, 5, 0],
[2, 1, 1, 1],
[1, 0, 0, 0],
]
)
}
print(raw_data)
print(raw_data[lane1.name].max(axis=1))
lanes = {lane1.name: lane1}
lanes[lane2.name] = lane2
norm_data = norm.AsIs().calc(lanes, raw_data)
plot_regions = mbf_genomics.regions.GenomicRegions(
"testregions", lambda: df, [], genome
)
with pytest.raises(KeyError):
o.calc(
plot_regions,
{lane1.name: lane1, lane2.name: lane2},
raw_data,
norm_data,
)
o = order.IthLaneMax(lane2)
with pytest.raises(KeyError):
o.calc(plot_regions, {lane1.name: lane1}, raw_data, norm_data)
raw_data[lane2.name] = raw_data[lane1.name].copy()
res_order, clusters = o.calc(plot_regions, lanes, raw_data, norm_data)
assert clusters is None
assert (
res_order == [2, 1, 0]
).all() # remember, from top to bottom in plotting later on.
raw_data[lane2.name] = np.array(
[
[0, 0, 0, 0],
[5, 1, 1, 0],
[1, 0, 0, 4],
]
)
o = order.IthLaneMax(0)
res_order, clusters = o.calc(plot_regions, lanes, raw_data, norm_data)
assert (
res_order == [2, 1, 0]
).all() # remember, from top to bottom in plotting later on.
o = order.IthLaneMax(1)
res_order, clusters = o.calc(plot_regions, lanes, raw_data, norm_data)
assert (
res_order == [0, 2, 1]
).all() # remember, from top to bottom in plotting later on.