Add fox lab to nwb

src/fox_lab_to_nwb/behavior.py

@@ -0,0 +1,107 @@
+from pathlib import Path
+from typing import Optional
+
+from neuroconv import BaseDataInterface
+from pydantic import FilePath
+from pynwb import NWBFile
+from pynwb import TimeSeries
+from pymatreader import read_mat
+
+
+class BehaviorInterface(BaseDataInterface):
+    def __init__(self, file_path: FilePath):
+        self.file_path = Path(file_path)
+
+    def add_to_nwbfile(
+        self,
+        nwbfile: NWBFile,
+        metadata: Optional[dict] = None,
+    ) -> None:
+
+        mat = read_mat(self.file_path)
+        rec_struct = mat["rec"]
+        daq_struct = rec_struct["daq"]
+
+        timestamps = daq_struct["tstamps"]
+
+        # Extracted from the matlab manually
+        # neither scipy or pymatreader can read string characters from the .mat file
+        channel_names = [
+            "CamSync",
+            "CamTrigger",
+            "OptoTrigger",
+            "LWingBeatAmp",
+            "RWingBeatAmp",
+            "WingBeatFreq",
+            "LHutchen",
+            "RHutchen",
+            "PTrigger",
+        ]
+
+        # TODO: figure how how to store this
+        # Synchronization signals
+        cam_sync = daq_struct["data"][:, 0]
+        cam_trigger = daq_struct["data"][:, 1]
+        opto_trigger = daq_struct["data"][:, 2]
+        ptrigger = daq_struct["data"][:, 8]
+
+        # Behavior signals
+        left_wing_beat_amplitude = daq_struct["data"][:, 3]
+        right_wing_beat_amplitude = daq_struct["data"][:, 4]
+        wing_beat_frequency = daq_struct["data"][:, 5]
+
+        unit = "tbd"
+        description = "tbd"
+        left_wing_beat_amplitude_time_series = TimeSeries(
+            name="LeftWingBeatAmplitudeTimeSeries",
+            data=left_wing_beat_amplitude,
+            unit=unit,
+            timestamps=timestamps,
+            description=description,
+        )
+
+        description = "tbd"
+        right_wing_beat_amplitude_time_series = TimeSeries(
+            name="RightWingBeatAmplitudeTimeSeries",
+            data=right_wing_beat_amplitude,
+            unit=unit,
+            timestamps=timestamps,
+            description=description,
+        )
+
+        description = "tbd"
+        unit = "Hz"  # TODO: Figure this out, the values in the plot are around 3 but should be higher for flies
+        wing_beat_frequency_time_series = TimeSeries(
+            name="WingBeatFrequencyTimeSeries",
+            data=wing_beat_frequency,
+            unit=unit,
+            timestamps=timestamps,
+            description=description,
+        )
+
+        nwbfile.add_acquisition(left_wing_beat_amplitude_time_series)
+        nwbfile.add_acquisition(right_wing_beat_amplitude_time_series)
+        nwbfile.add_acquisition(wing_beat_frequency_time_series)
+
+        # TODO: Ask Ben if this nesting makes sense? probably not
+        # time_series = [left_wing_beat_amplitude_time_series, right_wing_beat_amplitude_time_series, wing_beat_frequency_time_series]
+        # behavioral_time_series_container = BehavioralTimeSeries(name="BehavioralTimeSeries", time_series=time_series)
+        # nwbfile.add_acquisition(behavioral_time_series_container)
+
+        # Not clear what are those signals, haltere?
+        lhutchen = daq_struct["data"][:, 6]
+        rhutchen = daq_struct["data"][:, 7]
+
+        unit = "tbd"
+        description = "tbd"
+        lhutchen_time_series = TimeSeries(
+            name="LHutchenTimeSeries", data=lhutchen, unit=unit, timestamps=timestamps, description=description
+        )
+
+        description = "tbd"
+        rhutchen_time_series = TimeSeries(
+            name="RHutchenTimeSeries", data=rhutchen, unit=unit, timestamps=timestamps, description=description
+        )
+
+        nwbfile.add_acquisition(lhutchen_time_series)
+        nwbfile.add_acquisition(rhutchen_time_series)

src/fox_lab_to_nwb/conversion_notes.md

@@ -0,0 +1,153 @@
+
+# General comments
+Point-person: Amy Streets ([email protected]) and Kris Lea ([email protected])
+
+https://sites.google.com/site/cwrufoxlab/people
+
+## Trial structure
+
+Organized with folder hierarchy, with one folder for all trials of a single type/cross containing individual trial folders (ex. \Tshx18D07\Tshx18D07_240124_115923_f3_r1 where the cross is Tsh x 18D07 lines, the individual trial folder is “cross” _ “date” _ “time” _ “fly number” _ “trial repeat number”).
+
+
+# DeepLabCut output files
+config files missing but Video and DLC output files are available
+
+### Pickle (metadata)
+
+Here is an example output:
+
+```
+{'data': {'start': 1710281222.2714453,
+  'stop': 1710281243.7364442,
+  'run_duration': 21.464998960494995,
+  'Scorer': 'DLC_resnet50_antennatrackingMar11shuffle1_100000',
+  'DLC-model-config file': {'stride': 8.0,
+   'weigh_part_predictions': False,
+   'weigh_negatives': False,
+   'fg_fraction': 0.25,
+   'mean_pixel': [123.68, 116.779, 103.939],
+   'shuffle': True,
+   'snapshot_prefix': 'C:\\Users\\FoxLab-Whiskey\\Desktop\\antennatracking-AMS-2024-03-11\\dlc-models\\iteration-0\\antennatrackingMar11-trainset95shuffle1\\test\\snapshot',
+   'log_dir': 'log',
+   'global_scale': 0.8,
+   'location_refinement': True,
+   'locref_stdev': 7.2801,
+   'locref_loss_weight': 1.0,
+   'locref_huber_loss': True,
+   'optimizer': 'sgd',
+   'intermediate_supervision': False,
+   'intermediate_supervision_layer': 12,
+   'regularize': False,
+   'weight_decay': 0.0001,
+   'crop_pad': 0,
+   'scoremap_dir': 'test',
+   'batch_size': 8,
+   'dataset_type': 'imgaug',
+   'deterministic': False,
+   'mirror': False,
+   'pairwise_huber_loss': True,
+   'weigh_only_present_joints': False,
+   'partaffinityfield_predict': False,
+   'pairwise_predict': False,
+   'all_joints': [[0], [1], [2], [3]],
+   'all_joints_names': ['L_ant', 'R_ant', 'L_base', 'R_base'],
+   'dataset': 'training-datasets\\iteration-0\\UnaugmentedDataSet_antennatrackingMar11\\antennatracking_AMS95shuffle1.mat',
+   'init_weights': 'C:/Users/FoxLab-Whiskey/Desktop/antennatracking-AMS-2024-03-11\\dlc-models\\iteration-0\\antennatrackingMar11-trainset95shuffle1\\train\\snapshot-100000',
+   'net_type': 'resnet_50',
+   'num_joints': 4,
+   'num_outputs': 1},
+  'fps': 30.0,
+  'batch_size': 8,
+  'frame_dimensions': (480, 640),
+  'nframes': 4351,
+  'iteration (active-learning)': 0,
+  'training set fraction': 0.95,
+  'cropping': False,
+  'cropping_parameters': [0, 640, 0, 480]}}
+```
+
+# Static TIFF files for confocal imaging
+Not yet available
+
+# Synchronization signal with Spike2
+Not yet available
+
+# Intracellular electrophysiology data including the voltage trace and stimulus trace
+
+Not yet available
+
+# Files on data shared 
+
+### Tshx18D07_240124_115923_f3_r1.fly2
+This is supposed to be a struct with DAQ data and the wingbeat analysis. Can't open it with matlab online unless extension is changed to .mat
+
+Channel names:
+- CamSync
+- CamTrigger
+- OptoTrigger
+- LWingBeatAmp
+- RWingBeatAmp
+- WingBeatFreq
+- LHutchen  # What does Hutchen mean? Could be halteres?
+- RHutchen
+- PTrigger
+
+These can't be extracted with pyton from matlab because they are strings. Are they always the same order?
+
+## Cameras 
+
+### Fastec
+
+Files:
+* TOPCAM_000000.avi
+* SIDE_000000.avi
+
+Format is avi, metadata is in xml
+
+https://www.fastecimaging.com/
+
+TopCam has a DLC analysis for the following body parts:
+
+bodyparts: L_ant, R_ant, L_base, R_base,
+
+Sidecam does not have DLC analysis
+
+### Phantom
+Files:
+* XZ_1_186.mp4
+
+Format is mp4, metadata is in xml
+
+This camera has an associated DLC analysis for the following body parts:
+
+bodyparts: haltere
+
+### Photron
+
+Files:
+* Tshx18D07_240124_115923_f3_r1_down20.mp4 [This seems to be the same as topcam]
+
+
+Not sure about this. The `.mii` metadata is not available and there is this extra file:
+
+
+It also does not have a corresponding DLC analysis
+
+### Graphical example for the trials
+![example](./assets/camera_views.png)
+
+See the example, we have SideCam, TopCam and XZ_1_186.mp4
+
+The other vide is  Tshx18D07_240124_115923_f3_r1_down20.mp4 which I guess was downsampled.
+
+
+## `Tshx18D07_f3_r1_trial.Tsh`
+This should be trial data.
+Unable to read in matlab online unless extension is changed to .mat
+
+## Tshx18D07_240124_115923_f3_r1_down20_PROC.mat
+This has the analysis for the wingbeat redone.
+
+
+## Tshx18D07_ant_top_f3_r1_500ms_dvProject.mat
+Not described

src/fox_lab_to_nwb/metadata.yaml

@@ -0,0 +1,20 @@
+NWBFile:
+  keywords:
+    - Fly
+    - Wingbeat
+    - Behavior
+  related_publications:
+    - https://doi.org/great_publication or link to APA or MLA citation of the publication
+  session_description:
+    A rich text description of the experiment. Can also just be the abstract of the publication.
+  experiment_description: 'Task: Rapid serial visual presentation (RSVP).'
+  institution:  Case Western Reserve University
+  lab: Fox
+  experimenter:
+    - Streets, Amy
+    - Lea, Kriss
+  surgery: TBD
+Subject:
+  sex: M
+  species: Drosophila melanogaster  # TODO: Figure this one out
+  description: fly