Merge pull request #19 from catalystneuro/add_cell_registration

Add cell registration

src/cai_lab_to_nwb/zaki_2024/interfaces/__init__.py

@@ -4,3 +4,4 @@
 from .zaki_2024_sleep_classification_interface import Zaki2024SleepClassificationInterface
 from .miniscope_imaging_interface import MiniscopeImagingInterface
 from .zaki_2024_shock_stimuli_interface import Zaki2024ShockStimuliInterface
+from .zaki_2024_cell_registration_interface import Zaki2024CellRegistrationInterface

src/cai_lab_to_nwb/zaki_2024/interfaces/zaki_2024_cell_registration_interface.py

@@ -0,0 +1,72 @@
+"""Primary class for converting experiment-specific cell registration output."""
+
+from neuroconv.basedatainterface import BaseDataInterface
+from neuroconv.utils import DeepDict
+from typing import Optional
+from pathlib import Path
+from pynwb import NWBFile
+from hdmf.common.table import DynamicTable, VectorData
+
+import pandas as pd
+
+
+class Zaki2024CellRegistrationInterface(BaseDataInterface):
+    """Adds a table to store the output of CellReg."""
+
+    keywords = ["cross sessions cell registration"]
+
+    def __init__(self, file_paths: list[Path], verbose: bool = False):
+
+        self.verbose = verbose
+        self.file_paths = file_paths
+        super().__init__(file_paths=file_paths)
+
+    def get_metadata(self) -> DeepDict:
+        # Automatically retrieve as much metadata as possible from the source files available
+        metadata = super().get_metadata()
+
+        return metadata
+
+    def add_to_nwbfile(
+        self,
+        nwbfile: NWBFile,
+        subject_id: str,
+        stub_test: bool = False,
+        metadata: Optional[dict] = None,
+    ):
+        processing_module = nwbfile.create_processing_module(
+            name="cell_registration",
+            description="Processing module for cross session cell registration. "
+            "Cells recorded across sessions were cross-registered using a previously published open-source "
+            "cross-registration algorithm, CellReg, using the spatial correlations of nearby cells to "
+            "determine whether highly correlated footprints close in space are likely to be the same cell across sessions."
+            "Each offline recording was cross-registered with all the encoding and recall sessions, "
+            "but not with the other offline sessions because cross-registering between all sessions would lead to too many conflicts and, "
+            "therefore, to no cells cross-registered across all sessions."
+            "Each table represents the output of the cross-registration between one offline sessions and all the encoding and recall sessions. "
+            "A table maps the global ROI ids (row of the table) to the ROI ids in each of cross-registered session's plane segmentation.",
+        )
+
+        for file_path in self.file_paths:
+            offline_session_name = Path(file_path).stem.split(f"{subject_id}_")[-1]
+            name = offline_session_name + "vsConditioningSessions"
+            data = pd.read_csv(file_path)
+
+            columns = [
+                VectorData(
+                    name=col,
+                    description=f"ROI indexes of plane segmentation of session {col}",
+                    data=data[col].tolist()[:100] if stub_test else data[col].tolist(),
+                )
+                for col in data.columns
+            ]
+
+            dynamic_table = DynamicTable(
+                name=name,
+                description="Table maps the global ROI ids (row of the table) to the ROI ids in each of cross-registered session's plane segmentation."
+                "The column names refer to the cross-registered session's ids"
+                "The values -9999 indicates no correspondence. ",
+                columns=columns,
+            )
+
+            processing_module.add(dynamic_table)

src/cai_lab_to_nwb/zaki_2024/zaki_2024_convert_week_session.py

@@ -3,10 +3,9 @@
 import time
 from natsort import natsorted
 from pathlib import Path
+import warnings
 from typing import Union
-from datetime import datetime, timedelta
-import pandas as pd
-import json
+import re
 from neuroconv.utils import load_dict_from_file, dict_deep_update
 
 from zaki_2024_nwbconverter import Zaki2024NWBConverter
@@ -49,6 +48,23 @@ def session_to_nwb(
     )
     conversion_options.update(dict(MultiEDFSignals=dict(stub_test=stub_test)))
 
+    # Add Cross session cell registration
+    main_folder = data_dir_path / f"Ca_EEG_Calcium/{subject_id}/SpatialFootprints"
+    pattern = re.compile(r"^CellRegResults_OfflineDay(\d+)Session(\d+)$")
+
+    file_paths = []
+    for folder in main_folder.iterdir():
+        match = pattern.match(folder.name)
+        if folder.is_dir() and match:
+            offline_day, session_number = match.groups()
+            filename = f"CellRegResults_{subject_id}_OfflineDay{offline_day}Session{session_number}.csv"
+            csv_file = folder / filename
+            assert csv_file.is_file(), f"Expected file not found: {csv_file}"
+            file_paths.append(csv_file)
+
+    source_data.update(dict(CellRegistration=dict(file_paths=file_paths)))
+    conversion_options.update(dict(CellRegistration=dict(stub_test=stub_test, subject_id=subject_id)))
+
     converter = Zaki2024NWBConverter(source_data=source_data)
 
     # Add datetime to conversion

src/cai_lab_to_nwb/zaki_2024/zaki_2024_nwbconverter.py

@@ -1,11 +1,7 @@
 """Primary NWBConverter class for this dataset."""
 
-from copy import deepcopy
-
 from neuroconv import NWBConverter
-from neuroconv.utils import DeepDict
 from neuroconv.datainterfaces import VideoInterface
-from typing import Dict
 
 from interfaces import (
     MinianSegmentationInterface,
@@ -16,6 +12,7 @@
     MiniscopeImagingInterface,
     MinianMotionCorrectionInterface,
     Zaki2024ShockStimuliInterface,
+    Zaki2024CellRegistrationInterface,
 )
 
 
@@ -32,41 +29,5 @@ class Zaki2024NWBConverter(NWBConverter):
         FreezingBehavior=EzTrackFreezingBehaviorInterface,
         Video=VideoInterface,
         ShockStimuli=Zaki2024ShockStimuliInterface,
+        CellRegistration=Zaki2024CellRegistrationInterface,
     )
-
-
-"""
-    # TODO decide which datastream set the session start time
-    def get_metadata(self) -> DeepDict:
-        if "" not in self.data_interface_objects:
-            return super().get_metadata()
-
-        # Explicitly set session_start_time to ... start time
-        metadata = super().get_metadata()
-        session_start_time = self.data_interface_objects[""]
-        metadata["NWBFile"]["session_start_time"] = session_start_time
-
-        return metadata
-    
-    # TODO Add cell global_ids
-    def add_to_nwbfile(self, nwbfile: NWBFile, metadata, conversion_options: Optional[dict] = None) -> None:
-        super().add_to_nwbfile(nwbfile=nwbfile, metadata=metadata, conversion_options=conversion_options)
-
-        if "MinianSegmentation" in self.data_interface_objects:
-            global_roi_ids = get_global_ids_from_csv()
-            add_cell_registration(
-                nwbfile=nwbfile,
-                global_roi_ids=global_roi_ids,
-                plane_segmentation_name="PlaneSegmentation",
-            )
-
-    # TODO discuss time alignment with author
-    def temporally_align_data_interfaces(self):
-        aligned_starting_time = 0
-        if "MiniscopeImaging" in self.data_interface_classes:
-            miniscope_interface = self.data_interface_classes["MiniscopeImaging"]
-            miniscope_interface.set_aligned_starting_time(aligned_starting_time=aligned_starting_time)
-        if "MinianSegmentation" in self.data_interface_classes:
-            minian_interface = self.data_interface_classes["MinianSegmentation"]
-            minian_interface.set_aligned_starting_time(aligned_starting_time=aligned_starting_time)
-"""