def test_assert_melt():
for metric in ["precision_recall", "percent_strong", "grit"]:
result = metric_melt(
df=df,
features=features,
metadata_features=meta_features,
similarity_metric="pearson",
eval_metric=metric,
)
result = assign_replicates(similarity_melted_df=result,
replicate_groups=replicate_groups)
assert_melt(result, eval_metric=metric)
# Note, not all alternative dummy metrics are provided, since many require
# the same melted dataframe
if metric == "precision_recall":
dummy_metrics = ["percent_strong"]
elif metric == "percent_strong":
dummy_metrics = ["precision_recall", "grit"]
elif metric == "grit":
dummy_metrics = ["percent_strong"]
for dummy_metric in dummy_metrics:
with pytest.raises(AssertionError) as ve:
output = assert_melt(result, eval_metric=dummy_metric)
assert (
"Stop! The eval_metric provided in 'metric_melt()' is incorrect!"
in str(ve.value))
def test_assign_replicates():
replicate_groups = ["Metadata_broad_sample", "Metadata_mg_per_ml"]
result = assign_replicates(similarity_melted_df=similarity_melted_df,
replicate_groups=replicate_groups)
expected_cols = ["{x}_replicate".format(x=x)
for x in replicate_groups] + ["group_replicate"]
# Other functions expect columns to exist
assert all([x in result.columns.tolist() for x in expected_cols])
# Given the example data, we expect a certain number of pairwise replicates
expected_replicates = list(result.loc[:, expected_cols].sum().values)
assert expected_replicates == [1248, 408, 408]
# Try with a different number of replicate groups
replicate_groups = [
"Metadata_broad_sample",
"Metadata_mg_per_ml",
"Metadata_plate_map_name",
]
result = assign_replicates(similarity_melted_df=similarity_melted_df,
replicate_groups=replicate_groups)
expected_cols = ["{x}_replicate".format(x=x)
for x in replicate_groups] + ["group_replicate"]
# Other functions expect columns to exist
assert all([x in result.columns.tolist() for x in expected_cols])
# Given the example data, we expect a certain number of pairwise replicates
expected_replicates = list(result.loc[:, expected_cols].sum().values)
assert expected_replicates == [1248, 408, 73536, 408]
# This function will fail if a replicate column is given that doesn't belong
with pytest.raises(AssertionError) as ve:
replicate_groups = ["MISSING_COLUMN"]
result = assign_replicates(similarity_melted_df=similarity_melted_df,
replicate_groups=replicate_groups)
assert "replicate_group not found in melted dataframe columns" in str(
ve.value)
def test_calculate_precision_recall():
similarity_melted_df = metric_melt(
df=df,
features=features,
metadata_features=meta_features,
similarity_metric="pearson",
eval_metric="precision_recall",
)
replicate_groups = ["Metadata_broad_sample"]
result = assign_replicates(similarity_melted_df=similarity_melted_df,
replicate_groups=replicate_groups).sort_values(
by="similarity_metric", ascending=False)
pair_ids = set_pair_ids()
replicate_group_cols = [
"{x}{suf}".format(x=x, suf=pair_ids[list(pair_ids)[0]]["suffix"])
for x in replicate_groups
]
example_group = result.groupby(replicate_group_cols).get_group(
name=("BRD-A38592941-001-02-7"))
assert example_group.shape[
0] == 383 * 6 # number of pairwise comparisons per dose
# Assert that the similarity metrics are sorted
assert (example_group.similarity_metric.diff().dropna() > 0).sum() == 0
# Perform the calculation!
result = pd.DataFrame(calculate_precision_recall(example_group, k=10),
columns=["result"])
expected_result = {"k": 10, "precision": 0.4, "recall": 0.1333}
expected_result = pd.DataFrame(expected_result,
index=["result"]).transpose()
assert_frame_equal(result, expected_result, check_less_precise=True)
# Check that recall is 1 when k is maximized
result = pd.DataFrame(
calculate_precision_recall(example_group, k=example_group.shape[0]),
columns=["result"],
)
assert result.loc["recall", "result"] == 1
control_ids = ["Chr2", "Luc", "LacZ"]
# In[4]:
# Melt the input profiles to long format
similarity_melted_df = metric_melt(
df=cell_health_df,
features=features,
metadata_features=meta_features,
similarity_metric=similarity_metric,
eval_metric=operation,
)
similarity_melted_df = assign_replicates(
similarity_melted_df=similarity_melted_df,
replicate_groups=replicate_groups)
print(similarity_melted_df.shape)
similarity_melted_df.head()
# In[5]:
non_replicate_cor_95th = (
similarity_melted_df.query("not group_replicate").groupby(
"Metadata_cell_line_pair_a")["similarity_metric"].quantile(
0.95).reset_index().rename(
{"Metadata_cell_line_pair_a": "cell_line"}, axis="columns"))
# Output results
output_file = pathlib.Path(
df = pd.read_csv(plate_files[plate])
# Determine feature class
features = infer_cp_features(df)
meta_features = infer_cp_features(df, metadata=True)
# Calculate and process pairwise similarity matrix
audit_df = metric_melt(
df=df,
features=features,
metadata_features=meta_features,
similarity_metric="pearson",
eval_metric="replicate_reproducibility",
)
audit_df = assign_replicates(similarity_melted_df=audit_df,
replicate_groups=audit_cols)
# What is 95% of the non replicate null distribution
cutoff = audit_df.query(
"not group_replicate").similarity_metric.quantile(0.95)
# Calculate a single number for percent strong
percent_strong = evaluate(
profiles=df,
features=features,
meta_features=meta_features,
replicate_groups=audit_cols,
operation="replicate_reproducibility",
similarity_metric="pearson",
replicate_reproducibility_quantile=0.95,
)
def test_calculate_grit():
result = assign_replicates(
similarity_melted_df=similarity_melted_df,
replicate_groups=[replicate_id, group_id],
)
assert_melt(result, eval_metric="grit")
example_group = result.groupby(replicate_col_name).get_group(
name=("MTOR-2"))
# Perform the calculation!
grit_result = pd.DataFrame(
calculate_grit(example_group,
control_perts=control_perts,
column_id_info=column_id_info),
columns=["result"],
)
expected_result = {
"perturbation": "MTOR-2",
"group": "MTOR",
"grit": 1.55075
}
expected_result = pd.DataFrame(expected_result,
index=["result"]).transpose()
assert_frame_equal(grit_result, expected_result, check_less_precise=True)
# Calculate grit will not work with singleton perturbations
# (no other perts in same group)
example_group = result.groupby(replicate_col_name).get_group(
name=("AURKB-2"))
grit_result = pd.DataFrame(
calculate_grit(example_group,
control_perts=control_perts,
column_id_info=column_id_info),
columns=["result"],
)
expected_result = {
"perturbation": "AURKB-2",
"group": "AURKB",
"grit": np.nan
}
expected_result = pd.DataFrame(expected_result,
index=["result"]).transpose()
assert_frame_equal(grit_result, expected_result, check_less_precise=True)
# Calculate grit will not work with the full dataframe
with pytest.raises(AssertionError) as ae:
result = calculate_grit(
similarity_melted_df,
control_perts=control_perts,
column_id_info=column_id_info,
)
assert "grit is calculated for each perturbation independently" in str(
ae.value)
# Calculate grit will not work with when control barcodes are missing
with pytest.raises(AssertionError) as ae:
result = calculate_grit(
example_group,
control_perts=["DOES NOT EXIST", "THIS ONE NEITHER"],
column_id_info=column_id_info,
)
assert "Error! No control perturbations found." in str(ae.value)