def test_volcano_plot(self):
ppg.util.global_pipegraph.quiet = False
import mbf_sampledata
pasilla_data = pd.read_csv(
mbf_sampledata.get_sample_path(
"mbf_comparisons/pasillaCount_deseq2.tsv.gz"),
sep=" ",
)
# pasilla_data = pasilla_data.set_index('Gene')
pasilla_data.columns = [str(x) for x in pasilla_data.columns]
treated = [x for x in pasilla_data.columns if x.startswith("treated")]
untreated = [
x for x in pasilla_data.columns if x.startswith("untreated")
]
pasilla_data = DelayedDataFrame("pasilla", pasilla_data)
comp = Comparisons(pasilla_data, {
"treated": treated,
"untreated": untreated
}).a_vs_b("treated", "untreated", TTest())
comp.filter([("log2FC", "|>=", 2.0), ("FDR", "<=", 0.05)])
prune_qc(lambda job: "volcano" in job.job_id)
run_pipegraph()
qc_jobs = list(get_qc_jobs())
qc_jobs = [x for x in qc_jobs if not x._pruned]
print(qc_jobs)
assert len(qc_jobs) == 1
assert_image_equal(qc_jobs[0].filenames[0])
def test_subtraction_by_read(self):
from mbf_sampledata import get_human_22_fake_genome
genome = get_human_22_fake_genome()
lane = mbf_align.AlignedSample(
"test_lane",
get_sample_data(Path("mbf_align/rnaseq_spliced_chr22.bam")),
genome,
False,
"AA123",
) # index creation is automatic
lane2 = mbf_align.AlignedSample(
"test_lane2",
get_sample_data(Path("mbf_align/rnaseq_spliced_chr22.bam")),
genome,
False,
"AA124",
) # index creation is automatic
lane3 = mbf_align.AlignedSample(
"test_lane3",
get_sample_data(Path("mbf_align/chipseq_chr22.bam")),
genome,
False,
"AA123",
) # index creation is automatic
lane3_subset = mbf_align.AlignedSample(
"test_lane3_subset",
get_sample_data(Path("mbf_align/chipseq_chr22_subset.bam")),
genome,
False,
"AA123",
) # index creation is automatic
lane_empty = lane.post_process(
mbf_align.post_process.SubtractOtherLane(lane2), new_name="empty")
lane_full = lane.post_process(
mbf_align.post_process.SubtractOtherLane(lane3), new_name="full")
lane_some = lane3.post_process(
mbf_align.post_process.SubtractOtherLane(lane3_subset),
result_dir="results/aligned/shu",
)
qc_jobs = [
lane_some.post_processor_qc_jobs, lane_full.post_processor_qc_jobs
]
prune_qc(lambda job: job in qc_jobs)
ppg.run_pipegraph()
assert Path(lane_empty.get_bam_names()[1]).exists()
assert Path(lane_full.get_bam_names()[1]).exists()
assert lane_empty.mapped_reads() == 0
assert lane_full.mapped_reads() == lane.mapped_reads()
assert lane.mapped_reads() != 0
assert (lane_some.mapped_reads() == lane3.mapped_reads() -
lane3_subset.mapped_reads())
assert lane3_subset.mapped_reads(
) # make sure there was something to subtract
assert "shu" in lane_some.get_bam_names()[0]
assert_image_equal(qc_jobs[0].filenames[0], "_result_dir")
assert_image_equal(qc_jobs[0].filenames[0])
def test_assert_images_equal_inside_class(self):
assert_image_equal(
Path(__file__).parent / "base_images" / "test_qc" / "_" /
"test_assert_images_equal.png")
with pytest.raises(ValueError):
assert_image_equal(
Path(__file__).parent / "base_images" / "test_qc" / "_" /
"test_assert_images_equal.png",
"_b",
)
def _test_qc_plots(self, filename, remaining_job_count, chdir="."):
lane = self.prep_lane()
prune_qc(lambda job: filename in job.job_id)
not_pruned_count = sum([1 for x in get_qc_jobs() if not x._pruned])
assert not_pruned_count == remaining_job_count # plot cache, plot_table, plot
ppg.run_pipegraph()
if chdir == '..':
fn = lane.result_dir / chdir / filename
else:
fn = lane.result_dir / chdir / f"{lane.name}_{filename}"
assert_image_equal(fn, suffix="_" + filename)
def test_very_simple(self):
df = pd.DataFrame({
"a1": [0, 1, 2],
"a2": [0.5, 1.5, 2.5],
"b1": [2, 1, 0],
"b2": [2.5, 0.5, 1],
})
ddf = DelayedDataFrame("test", df)
of = "test.png"
h = HeatmapPlot(ddf, df.columns, of, heatmap_norm.Unchanged(),
heatmap_order.Unchanged())
run_pipegraph()
assert_image_equal(h.output_filename)
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_venn_from_logfcs(self):
ppg.util.global_pipegraph.quiet = False
d = DelayedDataFrame(
"ex1",
pd.DataFrame(
{
"gene_stable_id": ["A", "B", "C", "D", "E"],
"a": [1, 1, 1, 1, 1],
"b": [1, 2, 3, 4, 5],
"c": [1, 1, 3, 0.5, 0.75],
}
),
)
comp = Comparisons(d, {"a": ["a"], "b": ["b"], "c": ["c"]})
a = comp.all_vs_b("a", Log2FC())
selected = {name: x.filter([("log2FC", "|>=", 1)]) for name, x in a.items()}
plot_job = venn.plot_venn("test", selected)
ppg.run_pipegraph()
assert_image_equal(plot_job.filenames[0], "_down")
assert_image_equal(plot_job.filenames[1], "_up")
def test_hierarchical_pearson(self):
df = pd.DataFrame({
"a1": [0, 1, 2],
"a2": [0.5, 1.5, 2.5],
"b1": [2, 1, 0],
"b2": [0.5, 0.5, 1],
})
df = df.sample(200, replace=True, random_state=500)
np.random.seed(500)
df += np.random.normal(0, 1, df.shape)
ddf = DelayedDataFrame("test", df)
of = "test.png"
h = HeatmapPlot(
ddf,
df.columns,
of,
heatmap_norm.Unchanged(),
heatmap_order.HierarchicalPearson(),
)
run_pipegraph()
assert_image_equal(h.output_filename)
def test_ma_plot(self):
ppg.util.global_pipegraph.quiet = False
pasilla_data, treated, untreated = get_pasilla_data_subset()
import numpy
numpy.random.seed(500)
comp = Comparisons(pasilla_data, {
"treated": treated,
"untreated": untreated
}).a_vs_b("treated", "untreated", TTest(), laplace_offset=1)
comp.filter([
("log2FC", "|>=", 2.0),
# ('FDR', '<=', 0.05),
])
prune_qc(lambda job: "ma_plot" in job.job_id)
run_pipegraph()
qc_jobs = list(get_qc_jobs())
qc_jobs = [x for x in qc_jobs if not x._pruned]
assert len(qc_jobs) == 1
assert_image_equal(qc_jobs[0].filenames[0])
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_correlation(self):
ppg.util.global_pipegraph.quiet = False
import mbf_sampledata
pasilla_data = pd.read_csv(
mbf_sampledata.get_sample_path(
"mbf_comparisons/pasillaCount_deseq2.tsv.gz"),
sep=" ",
)
# pasilla_data = pasilla_data.set_index('Gene')
pasilla_data.columns = [str(x) for x in pasilla_data.columns]
treated = [x for x in pasilla_data.columns if x.startswith("treated")]
untreated = [
x for x in pasilla_data.columns if x.startswith("untreated")
]
pasilla_data = DelayedDataFrame("pasilla", pasilla_data)
Comparisons(pasilla_data, {"treated": treated, "untreated": untreated})
prune_qc(lambda job: "correlation" in job.job_id)
run_pipegraph()
qc_jobs = list(get_qc_jobs())
qc_jobs = [x for x in qc_jobs if not x._pruned]
print(qc_jobs)
assert len(qc_jobs) == 1
assert_image_equal(qc_jobs[0].filenames[0])
def test_assert_images_equal():
assert_image_equal(
Path(__file__).parent / "base_images" / "test_qc" / "_" /
"test_assert_images_equal.png")
with pytest.raises(
ValueError) as e: # here the baseline image does not exist
assert_image_equal(
Path(__file__).parent / "base_images" / "test_qc" / "_" /
"test_assert_images_equal.png",
"_b",
)
# should_path overwrites suffix
assert_image_equal(
Path(__file__).parent / "base_images" / "test_qc" / "_" /
"test_assert_images_equal.png",
"_b",
should_path=Path(__file__).parent / "base_images" / "test_qc" / "_" /
"test_assert_images_equal.png",
)
assert "Base_line image not found" in str(e.value)
with pytest.raises(ValueError) as e: # here it is different
assert_image_equal(
Path(__file__).parent / "base_images" / "test_qc" / "_" /
"test_assert_images_equal.png",
suffix="_c",
)
assert "Image files did not match" in str(e.value)
with pytest.raises(IOError) as e:
assert_image_equal("does not exist")
assert "not created" in str(e.value)
# with pytest.raises(ValueError) as e: #here it is different
with pytest.raises(ValueError) as e: # here it is different
assert_image_equal(
Path(__file__).parent / "base_images" / "test_qc" / "_" /
"test_assert_images_equal.png",
suffix="_d",
)
assert "do not match expected size" in str(e.value)