napistu 0.3.5__py3-none-any.whl → 0.3.7__py3-none-any.whl

tests/test_sbml_dfs_utils.py

@@ -3,7 +3,14 @@ from __future__ import annotations
 import pandas as pd
 
 from napistu import sbml_dfs_utils
-from napistu.constants import BQB, BQB_DEFINING_ATTRS, BQB_DEFINING_ATTRS_LOOSE
+from napistu.constants import (
+    BQB,
+    BQB_DEFINING_ATTRS,
+    BQB_DEFINING_ATTRS_LOOSE,
+    SBML_DFS,
+    IDENTIFIERS,
+    SBOTERM_NAMES,
+)
 
 
 def test_id_formatter():
@@ -17,47 +24,198 @@ def test_id_formatter():
     assert list(input_vals) == inv_ids
 
 
-def test_get_characteristic_species_ids():
-    """
-    Test get_characteristic_species_ids function with both dogmatic and non-dogmatic cases.
-    """
-    # Create mock species identifiers data
-    mock_species_ids = pd.DataFrame(
-        {
-            "s_id": ["s1", "s2", "s3", "s4", "s5"],
-            "identifier": ["P12345", "CHEBI:15377", "GO:12345", "P67890", "P67890"],
-            "ontology": ["uniprot", "chebi", "go", "uniprot", "chebi"],
-            "bqb": [
-                "BQB_IS",
-                "BQB_IS",
-                "BQB_HAS_PART",
-                "BQB_HAS_VERSION",
-                "BQB_ENCODES",
-            ],
-        }
+def test_filter_to_characteristic_species_ids():
+
+    species_ids_dict = {
+        SBML_DFS.S_ID: ["large_complex"] * 6
+        + ["small_complex"] * 2
+        + ["proteinA", "proteinB"]
+        + ["proteinC"] * 3
+        + [
+            "promiscuous_complexA",
+            "promiscuous_complexB",
+            "promiscuous_complexC",
+            "promiscuous_complexD",
+            "promiscuous_complexE",
+        ],
+        IDENTIFIERS.ONTOLOGY: ["complexportal"]
+        + ["HGNC"] * 7
+        + ["GO"] * 2
+        + ["ENSG", "ENSP", "pubmed"]
+        + ["HGNC"] * 5,
+        IDENTIFIERS.IDENTIFIER: [
+            "CPX-BIG",
+            "mem1",
+            "mem2",
+            "mem3",
+            "mem4",
+            "mem5",
+            "part1",
+            "part2",
+            "GO:1",
+            "GO:2",
+            "dna_seq",
+            "protein_seq",
+            "my_cool_pub",
+        ]
+        + ["promiscuous_complex"] * 5,
+        IDENTIFIERS.BQB: [BQB.IS]
+        + [BQB.HAS_PART] * 7
+        + [BQB.IS] * 2
+        + [
+            # these are retained if BQB_DEFINING_ATTRS_LOOSE is used
+            BQB.ENCODES,
+            BQB.IS_ENCODED_BY,
+            # this should always be removed
+            BQB.IS_DESCRIBED_BY,
+        ]
+        + [BQB.HAS_PART] * 5,
+    }
+
+    species_ids = pd.DataFrame(species_ids_dict)
+
+    characteristic_ids_narrow = sbml_dfs_utils.filter_to_characteristic_species_ids(
+        species_ids,
+        defining_biological_qualifiers=BQB_DEFINING_ATTRS,
+        max_complex_size=4,
+        max_promiscuity=4,
+    )
+
+    EXPECTED_IDS = ["CPX-BIG", "GO:1", "GO:2", "part1", "part2"]
+    assert characteristic_ids_narrow[IDENTIFIERS.IDENTIFIER].tolist() == EXPECTED_IDS
+
+    characteristic_ids_loose = sbml_dfs_utils.filter_to_characteristic_species_ids(
+        species_ids,
+        # include encodes and is_encoded_by as equivalent to is
+        defining_biological_qualifiers=BQB_DEFINING_ATTRS_LOOSE,
+        max_complex_size=4,
+        # expand promiscuity to default value
+        max_promiscuity=20,
     )
 
-    # Create mock SBML_dfs object
-    class MockSBML_dfs:
-        def get_identifiers(self, entity_type):
-            return mock_species_ids
+    EXPECTED_IDS = [
+        "CPX-BIG",
+        "GO:1",
+        "GO:2",
+        "dna_seq",
+        "protein_seq",
+        "part1",
+        "part2",
+    ] + ["promiscuous_complex"] * 5
+    assert characteristic_ids_loose[IDENTIFIERS.IDENTIFIER].tolist() == EXPECTED_IDS
 
-    mock_sbml = MockSBML_dfs()
 
-    # Test dogmatic case (default)
-    expected_bqbs = BQB_DEFINING_ATTRS + [BQB.HAS_PART]  # noqa: F841
-    dogmatic_result = sbml_dfs_utils.get_characteristic_species_ids(mock_sbml)
-    expected_dogmatic = mock_species_ids.query("bqb in @expected_bqbs")
+def test_formula(sbml_dfs):
+    # create a formula string
 
-    pd.testing.assert_frame_equal(dogmatic_result, expected_dogmatic, check_like=True)
+    an_r_id = sbml_dfs.reactions.index[0]
 
-    # Test non-dogmatic case
-    expected_bqbs = BQB_DEFINING_ATTRS_LOOSE + [BQB.HAS_PART]  # noqa: F841
-    non_dogmatic_result = sbml_dfs_utils.get_characteristic_species_ids(
-        mock_sbml, dogmatic=False
+    reaction_species_df = sbml_dfs.reaction_species[
+        sbml_dfs.reaction_species["r_id"] == an_r_id
+    ].merge(sbml_dfs.compartmentalized_species, left_on="sc_id", right_index=True)
+
+    formula_str = sbml_dfs_utils.construct_formula_string(
+        reaction_species_df, sbml_dfs.reactions, name_var="sc_name"
     )
-    expected_non_dogmatic = mock_species_ids.query("bqb in @expected_bqbs")
 
-    pd.testing.assert_frame_equal(
-        non_dogmatic_result, expected_non_dogmatic, check_like=True
+    assert isinstance(formula_str, str)
+    assert (
+        formula_str
+        == "CO2 [extracellular region] -> CO2 [cytosol] ---- modifiers: AQP1 tetramer [plasma membrane]]"
     )
+
+
+def test_find_underspecified_reactions():
+
+    reaction_w_regulators = pd.DataFrame(
+        {
+            SBML_DFS.SC_ID: ["A", "B", "C", "D", "E", "F", "G"],
+            SBML_DFS.STOICHIOMETRY: [-1, -1, 1, 1, 0, 0, 0],
+            SBML_DFS.SBO_TERM: [
+                SBOTERM_NAMES.REACTANT,
+                SBOTERM_NAMES.REACTANT,
+                SBOTERM_NAMES.PRODUCT,
+                SBOTERM_NAMES.PRODUCT,
+                SBOTERM_NAMES.CATALYST,
+                SBOTERM_NAMES.CATALYST,
+                SBOTERM_NAMES.STIMULATOR,
+            ],
+        }
+    ).assign(r_id="bar")
+    reaction_w_regulators[SBML_DFS.RSC_ID] = [
+        f"rsc_{i}" for i in range(len(reaction_w_regulators))
+    ]
+    reaction_w_regulators.set_index(SBML_DFS.RSC_ID, inplace=True)
+    reaction_w_regulators = sbml_dfs_utils.add_sbo_role(reaction_w_regulators)
+
+    reaction_w_interactors = pd.DataFrame(
+        {
+            SBML_DFS.SC_ID: ["A", "B"],
+            SBML_DFS.STOICHIOMETRY: [-1, 1],
+            SBML_DFS.SBO_TERM: [SBOTERM_NAMES.REACTANT, SBOTERM_NAMES.REACTANT],
+        }
+    ).assign(r_id="baz")
+    reaction_w_interactors[SBML_DFS.RSC_ID] = [
+        f"rsc_{i}" for i in range(len(reaction_w_interactors))
+    ]
+    reaction_w_interactors.set_index(SBML_DFS.RSC_ID, inplace=True)
+    reaction_w_interactors = sbml_dfs_utils.add_sbo_role(reaction_w_interactors)
+
+    working_reactions = reaction_w_regulators.copy()
+    working_reactions["new"] = True
+    working_reactions.loc["rsc_0", "new"] = False
+    working_reactions
+    result = sbml_dfs_utils._find_underspecified_reactions(working_reactions)
+    assert result == {"bar"}
+
+    # missing one enzyme -> operable
+    working_reactions = reaction_w_regulators.copy()
+    working_reactions["new"] = True
+    working_reactions.loc["rsc_4", "new"] = False
+    working_reactions
+    result = sbml_dfs_utils._find_underspecified_reactions(working_reactions)
+    assert result == set()
+
+    # missing one product -> inoperable
+    working_reactions = reaction_w_regulators.copy()
+    working_reactions["new"] = True
+    working_reactions.loc["rsc_2", "new"] = False
+    working_reactions
+    result = sbml_dfs_utils._find_underspecified_reactions(working_reactions)
+    assert result == {"bar"}
+
+    # missing all enzymes -> inoperable
+    working_reactions = reaction_w_regulators.copy()
+    working_reactions["new"] = True
+    working_reactions.loc["rsc_4", "new"] = False
+    working_reactions.loc["rsc_5", "new"] = False
+    working_reactions
+    result = sbml_dfs_utils._find_underspecified_reactions(working_reactions)
+    assert result == {"bar"}
+
+    # missing regulators -> operable
+    working_reactions = reaction_w_regulators.copy()
+    working_reactions["new"] = True
+    working_reactions.loc["rsc_6", "new"] = False
+    working_reactions
+    result = sbml_dfs_utils._find_underspecified_reactions(working_reactions)
+    assert result == set()
+
+    # remove an interactor
+    working_reactions = reaction_w_interactors.copy()
+    working_reactions["new"] = True
+    working_reactions.loc["rsc_0", "new"] = False
+    working_reactions
+    result = sbml_dfs_utils._find_underspecified_reactions(working_reactions)
+    assert result == {"baz"}
+
+
+def test_stubbed_compartment():
+    compartment = sbml_dfs_utils.stub_compartments()
+
+    assert compartment["c_Identifiers"].iloc[0].ids[0] == {
+        "ontology": "go",
+        "identifier": "GO:0005575",
+        "url": "https://www.ebi.ac.uk/QuickGO/term/GO:0005575",
+        "bqb": "BQB_IS",
+    }

tests/test_utils.py

@@ -686,3 +686,22 @@ def test_safe_fill():
         "a_very_long\nstringggg",
         "",
     ]
+
+
+def test_update_pathological_names():
+
+    # All numeric
+    s = pd.Series(["1", "2", "3"])
+    out = utils.update_pathological_names(s, "prefix_")
+    assert all(x.startswith("prefix_") for x in out)
+    assert list(out) == ["prefix_1", "prefix_2", "prefix_3"]
+
+    # Mixed numeric and non-numeric
+    s2 = pd.Series(["1", "foo", "3"])
+    out2 = utils.update_pathological_names(s2, "prefix_")
+    assert list(out2) == ["1", "foo", "3"]
+
+    # All non-numeric
+    s3 = pd.Series(["foo", "bar", "baz"])
+    out3 = utils.update_pathological_names(s3, "prefix_")
+    assert list(out3) == ["foo", "bar", "baz"]