Add subject metadata from excel sheet

src/howe_lab_to_nwb/vu2024/vu2024_convert_all_sessions.py

@@ -1,4 +1,5 @@
 import re
+from datetime import datetime
 from pathlib import Path
 
 import pandas as pd
@@ -40,7 +41,10 @@ def convert_all_sessions(
 
     """
 
-    data_table = pd.read_excel(data_table_path)
+    data_table_dict = pd.read_excel(data_table_path, sheet_name=["Sessions", "Mice"])
+    data_table = data_table_dict["Sessions"]
+    subjects_table = data_table_dict["Mice"].astype(str)
+
     folder_path = Path(folder_path)
 
     columns_to_group = ["Mouse", "Date (YYYYMMDD)"]
@@ -55,8 +59,22 @@ def convert_all_sessions(
 
     for (subject_id, date), table in progress_bar:
         experiment_type = "single-wavelength" if len(table) == 1 else "dual-wavelength"
-        subject_id = str(subject_id).replace("-", "")
 
+        mouse_metadata = subjects_table.loc[subjects_table["MouseID"] == str(subject_id)]
+        mouse_metadata = mouse_metadata.to_dict(orient="records")[0]
+        date_of_birth = datetime.strptime(mouse_metadata["Date of Birth"], "%Y-%m-%d")
+        subject_metadata = dict(
+            Subject=dict(
+                subject_id=mouse_metadata["MouseID"],
+                date_of_birth=date_of_birth,
+                sex=mouse_metadata["Sex"],
+                genotype=mouse_metadata["Genotype"],
+                strain=mouse_metadata["Strain"],
+                species="Mus musculus",
+            ),
+        )
+
+        subject_id = str(subject_id).replace("-", "")
         subject_folder_path = folder_path / subject_id
 
         behavior_file_name = table["Raw behavior file"].values[0]
@@ -143,6 +161,7 @@ def convert_all_sessions(
                 ttl_stream_names=ttl_stream_names,
                 behavior_file_paths=behavior_file_paths,
                 nwbfile_path=nwbfile_path,
+                subject_metadata=subject_metadata,
                 stub_test=stub_test,
             )
         elif experiment_type == "single-wavelength":
@@ -157,6 +176,7 @@ def convert_all_sessions(
                 ttl_stream_name=ttl_stream_names[0],
                 behavior_file_path=behavior_file_paths[0],
                 nwbfile_path=nwbfile_path,
+                subject_metadata=subject_metadata,
                 stub_test=stub_test,
             )
         progress_bar.update(1)
@@ -186,12 +206,12 @@ def convert_all_sessions(
     nwbfile_folder_path = Path("/Volumes/t7-ssd/Howe/nwbfiles")
 
     # Whether to overwrite existing NWB files, default is False
-    overwrite = False
+    overwrite = True
 
     # Whether to run the conversion as a stub test
     # When set to True, write only a subset of the data for each session
     # When set to False, write the entire data for each session
-    stub_test = False
+    stub_test = True
 
     convert_all_sessions(
         data_table_path=data_table_excel_file_path,

src/howe_lab_to_nwb/vu2024/vu2024_convert_dual_wavelength_session.py

@@ -19,6 +19,7 @@ def dual_wavelength_session_to_nwb(
     nwbfile_path: Union[str, Path],
     behavior_file_paths: List[Union[str, Path]] = None,
     motion_corrected_imaging_file_paths: List[Union[str, Path]] = None,
+    subject_metadata: dict = None,
     stub_test: bool = False,
 ):
     """
@@ -46,6 +47,8 @@ def dual_wavelength_session_to_nwb(
         The list of paths to the .mat files containing the processed behavior data.
     motion_corrected_imaging_file_paths : List[Union[str, Path]], optional
         The list of paths to the .tif files containing the motion corrected imaging data.
+    subject_metadata : dict, optional
+        The metadata for the subject.
     stub_test : bool, optional
         Whether to run the conversion as a stub test.
     """
@@ -77,6 +80,7 @@ def dual_wavelength_session_to_nwb(
         ttl_stream_name=ttl_stream_names[0],
         motion_corrected_imaging_file_path=motion_corrected_imaging_file_paths[0],
         behavior_file_path=behavior_file_paths[0],
+        subject_metadata=subject_metadata,
         stub_test=stub_test,
     )
 

src/howe_lab_to_nwb/vu2024/vu2024_convert_single_wavelength_session.py

@@ -26,6 +26,7 @@ def single_wavelength_session_to_nwb(
     nwbfile_path: Optional[Union[str, Path]] = None,
     nwbfile: Optional[NWBFile] = None,
     sampling_frequency: float = None,
+    subject_metadata: Optional[dict] = None,
     stub_test: bool = False,
 ) -> NWBFile:
     """
@@ -64,6 +65,8 @@ def single_wavelength_session_to_nwb(
     sampling_frequency : float, optional
         The sampling frequency of the data. If None, the sampling frequency will be read from the .cxd file.
         If missing from the file, the sampling frequency must be provided.
+    subject_metadata : dict, optional
+        The metadata for the subject.
     stub_test : bool, optional
         Whether to run a stub test, by default False.
     """
@@ -157,6 +160,11 @@ def single_wavelength_session_to_nwb(
     editable_metadata = load_dict_from_file(editable_metadata_path)
     metadata = dict_deep_update(metadata, editable_metadata)
 
+    if subject_metadata is not None:
+        metadata = dict_deep_update(metadata, subject_metadata)
+        date_of_birth = metadata["Subject"]["date_of_birth"]
+        metadata["Subject"].update(date_of_birth=date_of_birth.replace(tzinfo=tzinfo))
+
     ophys_metadata = load_dict_from_file(Path(__file__).parent / "metadata" / "vu2024_ophys_metadata.yaml")
     metadata = dict_deep_update(metadata, ophys_metadata)