Update CellLineMetaData (#334)

* update cell line annotation * add LookUp object * pass nox test * fix test * keep downloaded files * improve error message and lookup initialize * add annotate_from_gdsc * add metadata source in the docstring * fix typo * update test script and speed up downloading metadata * add remotezip dependency * parametrize Pytest fixture * fix import error * poetry update Signed-off-by: zethson <[email protected]> * Fix fixture Signed-off-by: zethson <[email protected]> * fix test error * Refactor Signed-off-by: zethson <[email protected]> * Refactor Signed-off-by: zethson <[email protected]> * Refactor Signed-off-by: zethson <[email protected]> * improve LookUp object * add openpyxl Signed-off-by: zethson <[email protected]> * update lookup * replace dict with namedtuple --------- Signed-off-by: zethson <[email protected]> Co-authored-by: Lukas Heumos <[email protected]>
scverse · Aug 8, 2023 · bbdf99b · bbdf99b
1 parent 5dc2b9b
commit bbdf99b
Show file tree

Hide file tree

Showing 8 changed files with 873 additions and 2,282 deletions.
diff --git a/pertpy/tools/_metadata/_cell_line.py b/pertpy/tools/_metadata/_cell_line.py
diff --git a/pertpy/tools/_metadata/_look_up.py b/pertpy/tools/_metadata/_look_up.py
diff --git a/pertpy/tools/_perturbation_space/_perturbation_space.py b/pertpy/tools/_perturbation_space/_perturbation_space.py
@@ -184,14 +184,14 @@ def add(
             for key in data["layers"]:
                 key_name = key
                 if key.endswith("_control_diff"):
-                    key_name = key.remove_suffix("_control_diff")
+                    key_name = key.removesuffix("_control_diff")
                 new_perturbation.layers[key_name] = data["layers"][key]
 
         if "embeddings" in data.keys():
             key_name = key
             for key in data["embeddings"]:
                 if key.endswith("_control_diff"):
-                    key_name = key.remove_suffix("_control_diff")
+                    key_name = key.removesuffix("_control_diff")
                 new_perturbation.obsm[key_name] = data["embeddings"][key]
 
         if ensure_consistency:
@@ -275,14 +275,14 @@ def subtract(
             for key in data["layers"]:
                 key_name = key
                 if key.endswith("_control_diff"):
-                    key_name = key.remove_suffix("_control_diff")
+                    key_name = key.removesuffix("_control_diff")
                 new_perturbation.layers[key_name] = data["layers"][key]
 
         if "embeddings" in data.keys():
             key_name = key
             for key in data["embeddings"]:
                 if key.endswith("_control_diff"):
-                    key_name = key.remove_suffix("_control_diff")
+                    key_name = key.removesuffix("_control_diff")
                 new_perturbation.obsm[key_name] = data["embeddings"][key]
 
         if ensure_consistency:

diff --git a/poetry.lock b/poetry.lock
diff --git a/pyproject.toml b/pyproject.toml
@@ -41,6 +41,8 @@ pyqt5 = {version = "^5.15.9", optional = true}
 ott-jax = "^0.4.0"
 sparsecca = ">=0.3.0"
 numba = "^0.57.1"
+remotezip = "^0.12.1"
+openpyxl = "^3.1.2"
 
 [tool.poetry.dev-dependencies]
 black = ">=22.12.0"

diff --git a/tests/tools/_metadata/test_cell_line.py b/tests/tools/_metadata/test_cell_line.py
@@ -1,6 +1,7 @@
 import anndata
 import numpy as np
 import pandas as pd
+import pytest
 from anndata import AnnData
 from scipy import sparse
 
@@ -12,7 +13,10 @@
 
 
 class TestMetaData:
-    def make_test_adata(self) -> AnnData:
+    pt_metadata = pt.tl.CellLineMetaData()
+
+    @pytest.fixture
+    def adata(self) -> AnnData:
         np.random.seed(1)
 
         X = np.random.normal(0, 1, (NUM_CELLS, NUM_GENES))
@@ -21,13 +25,14 @@ def make_test_adata(self) -> AnnData:
         cell_line = {
             "DepMap_ID": ["ACH-000016"] * NUM_CELLS_PER_ID
             + ["ACH-000049"] * NUM_CELLS_PER_ID
-            + ["ACH-000130"] * NUM_CELLS_PER_ID
-            + ["ACH-000216"] * NUM_CELLS_PER_ID
+            + ["ACH-001208"] * NUM_CELLS_PER_ID
+            + ["ACH-000956"] * NUM_CELLS_PER_ID
         }
         cell_line = pd.DataFrame(cell_line)
         obs = pd.concat([cell_line], axis=1)
         obs = obs.set_index(np.arange(NUM_GENES))
         obs.index.rename("index", inplace=True)
+        obs["perturbation"] = "Midostaurin"
 
         var_data = {"gene_name": ["gene" + str(i) for i in range(1, NUM_GENES + 1)]}
         var = pd.DataFrame(var_data)
@@ -39,59 +44,51 @@ def make_test_adata(self) -> AnnData:
 
         return adata
 
-    def test_cell_line_annotation(self):
-        adata = self.make_test_adata()
-        pt_metadata = pt.tl.CellLineMetaData()
-        pt_metadata.annotate_cell_lines(adata=adata)
-
-        assert len(adata.obs.columns) == len(pt_metadata.cell_line_meta.columns)
-        assert set(pt_metadata.cell_line_meta.columns).issubset(adata.obs)
+    def test_cell_line_annotation(self, adata):
+        self.pt_metadata.annotate_cell_lines(adata=adata)
+        assert (
+            len(adata.obs.columns) == len(self.pt_metadata.cell_line_meta.columns) + 1
+        )  # due to the perturbation column
+        assert set(self.pt_metadata.cell_line_meta.columns).issubset(adata.obs)
         stripped_cell_line_name = (
             ["SLR21"] * NUM_CELLS_PER_ID
             + ["HEKTE"] * NUM_CELLS_PER_ID
-            + ["NALM19"] * NUM_CELLS_PER_ID
-            + ["JHESOAD1"] * NUM_CELLS_PER_ID
+            + ["TK10"] * NUM_CELLS_PER_ID
+            + ["22RV1"] * NUM_CELLS_PER_ID
         )
 
         assert stripped_cell_line_name == list(adata.obs["stripped_cell_line_name"])
 
-    def test_ccle_expression_annotation(self):
-        adata = self.make_test_adata()
-        pt_metadata = pt.tl.CellLineMetaData()
-        pt_metadata.annotate_cell_lines(adata)
-        pt_metadata.annotate_ccle_expression(adata)
-
-        assert len(adata.obsm) == 1
-        assert adata.obsm["CCLE_expression"].shape == (NUM_CELLS, len(pt_metadata.ccle_expr.columns))
+    def test_gdsc_annotation(self, adata):
+        self.pt_metadata.annotate_cell_lines(adata)
+        self.pt_metadata.annotate_from_gdsc(adata, query_id="stripped_cell_line_name")
+        assert "drug_name" in adata.obs
+        assert "ln_ic50" in adata.obs
 
-    def test_protein_expression_annotation(self):
-        adata = self.make_test_adata()
-        pt_metadata = pt.tl.CellLineMetaData()
-        pt_metadata.annotate_cell_lines(adata)
-        pt_metadata.annotate_protein_expression(adata)
+    def test_protein_expression_annotation(self, adata):
+        self.pt_metadata.annotate_cell_lines(adata)
+        self.pt_metadata.annotate_protein_expression(adata, query_id="stripped_cell_line_name")
 
         assert len(adata.obsm) == 1
         assert adata.obsm["proteomics_protein_intensity"].shape == (
             NUM_GENES,
-            len(pt_metadata.proteomics_data.uniprot_id.unique()),
+            len(self.pt_metadata.proteomics_data.uniprot_id.unique()),
         )
 
-    def test_bulk_rna_expression_annotation(self):
-        adata = self.make_test_adata()
-        pt_metadata = pt.tl.CellLineMetaData()
-        pt_metadata.annotate_cell_lines(adata)
-        pt_metadata.annotate_bulk_rna_expression(adata)
+    def test_bulk_rna_expression_annotation(self, adata):
+        self.pt_metadata.annotate_cell_lines(adata)
+        self.pt_metadata.annotate_bulk_rna_expression(adata, query_id="DepMap_ID", cell_line_source="broad")
 
         assert len(adata.obsm) == 1
         assert adata.obsm["bulk_rna_expression_broad"].shape == (
             NUM_GENES,
-            len(pt_metadata.bulk_rna_broad.gene_id.unique()),
+            self.pt_metadata.bulk_rna_broad.shape[1],
         )
 
-        pt_metadata.annotate_bulk_rna_expression(adata, bulk_rna_source="sanger")
+        self.pt_metadata.annotate_bulk_rna_expression(adata, query_id="stripped_cell_line_name")
 
         assert len(adata.obsm) == 2
         assert adata.obsm["bulk_rna_expression_sanger"].shape == (
             NUM_GENES,
-            len(pt_metadata.bulk_rna_sanger.gene_id.unique()),
+            self.pt_metadata.bulk_rna_sanger.shape[1],
         )