Merge branch 'master' into group_abundance

Former-commit-id: 451eb4b9338e477f18944867fba3b231f8714d79

.gitignore

@@ -1,5 +1,5 @@
-.vscode
 notebooks
+tutorial/TraCeR
 # Byte-compiled / optimized / DLL files
 __pycache__/
 *.py[cod]

.vscode/settings.json

@@ -0,0 +1,4 @@
+{
+  "restructuredtext.confPath": "${workspaceFolder}/docs",
+  "editor.rulers": [88]
+}

sctcrpy/_io/_io.py

@@ -287,15 +287,20 @@ def _process_chains(chains, chain_type):
     ):
         cell_name = summary_file.split(os.sep)[-3]
         tcr_obj = TcrCell(cell_name)
-        with open(summary_file, "rb") as f:
-            tracer_obj = pickle.load(f)
-            chains = tracer_obj.recombinants["TCR"]
-            if "A" in chains:
-                for tmp_chain in _process_chains(chains["A"], "TRA"):
-                    tcr_obj.add_chain(tmp_chain)
-            if "B" in chains:
-                for tmp_chain in _process_chains(chains["B"], "TRB"):
-                    tcr_obj.add_chain(tmp_chain)
+        try:
+            with open(summary_file, "rb") as f:
+                tracer_obj = pickle.load(f)
+                chains = tracer_obj.recombinants["TCR"]
+                if "A" in chains and chains["A"] is not None:
+                    for tmp_chain in _process_chains(chains["A"], "TRA"):
+                        tcr_obj.add_chain(tmp_chain)
+                if "B" in chains and chains["B"] is not None:
+                    for tmp_chain in _process_chains(chains["B"], "TRB"):
+                        tcr_obj.add_chain(tmp_chain)
+        except Exception as e:
+            raise Exception(
+                "Error loading TCR data from cell {}".format(summary_file)
+            ) from e
 
         tcr_objs[cell_name] = tcr_obj
 

sctcrpy/_tools/_tools.py

@@ -7,7 +7,11 @@
 
 
 def define_clonotypes(
-    adata: AnnData, *, inplace: bool = True, key_added: str = "clonotype"
+    adata: AnnData,
+    *,
+    flavor: Literal["all_chains", "primary_only"] = "all_chains",
+    inplace: bool = True,
+    key_added: str = "clonotype",
 ) -> Union[None, np.ndarray]:
     """Define clonotypes based on CDR3 region.
 
@@ -20,6 +24,10 @@ def define_clonotypes(
     ----------
     adata
         Annotated data matrix
+    flavor
+        Biological model to define clonotypes. 
+        `all_chains`: All four chains of a cell in a clonotype need to be the same. 
+        `primary_only`: Only primary alpha and beta chain need to be the same. 
     inplace
         If True, adds a column to adata.obs
     key_added
@@ -32,12 +40,14 @@ def define_clonotypes(
     or adds a `clonotype` column to `adata`. 
     
     """
+    groupby_cols = {
+        "all_chains": ["TRA_1_cdr3", "TRB_1_cdr3", "TRA_2_cdr3", "TRA_2_cdr3"],
+        "primary_only": ["TRA_1_cdr3", "TRB_1_cdr3"],
+    }
     clonotype_col = np.array(
         [
             "clonotype_{}".format(x)
-            for x in adata.obs.groupby(
-                ["TRA_1_cdr3", "TRB_1_cdr3", "TRA_2_cdr3", "TRA_2_cdr3"]
-            ).ngroup()
+            for x in adata.obs.groupby(groupby_cols[flavor]).ngroup()
         ]
     )
     clonotype_col[

tests/test_tools.py

@@ -28,6 +28,16 @@ def test_define_clonotypes():
         ["clonotype_1", np.nan, "clonotype_1", "clonotype_0", "clonotype_2"],
     )
 
+    res_primary_only = st.tl.define_clonotypes(
+        adata, flavor="primary_only", inplace=False
+    )
+    npt.assert_equal(
+        # order is by alphabet: BBBB < nan
+        # we don't care about the order of numbers, so this is ok.
+        res_primary_only,
+        ["clonotype_0", np.nan, "clonotype_0", "clonotype_0", "clonotype_1"],
+    )
+
     # test inplace
     st.tl.define_clonotypes(adata, key_added="clonotype_")
     npt.assert_equal(res, adata.obs["clonotype_"].values)