Howe Lab Conversion

[weiglszonja]

Fiber photometry and Imaging (Vu 2024)

• collect initial notes: Vu2024 - Fiber photometry conversion notes #2
• ask for missing data
• Add extractor and interface for Imaging (Hamamatsu .cdx file) (Add CXD imaging #5): integration with NeuroConv progress: Add extractor for BioFormats (CXD) roiextractors#338
• Add motion corrected imaging data (_MC.tif file) Add motion correction from TIF #6
• Add the extracted fluorescence time series data (___ROIs.mat) in progress Add raw fiber photometry #4
• Annotate fibers (___fiber_locations.xlsx)
• Add behavioral camera recordings as external file (Add behavior signals (velocity and binary: lick, light, tone, reward) and camera recording #7)
• Add raw behavior file (___YYYY.MM.DD_hh.mm.ss.mat) -> using this to set the timestamps for imaging and fiber photometry from the TTL signal
• Add preprocessed behavior file (___YYYY.MM.DD_hh.mm.ss_ttlInX_movieX.mat): the raw behavior data has been downsampled via block averaging to match the sampling rate of the neural data Add behavior signals (velocity and binary: lick, light, tone, reward) and camera recording #7
• Add ROI masks and locations in a PlaneSegmentation table Add custom segmentation #8
• Enable multi-camera recordings Add support for multiple videos #10
• Enable dual-wavelength excitation recordings Add dual-wavelength sessions #9
• Run conversion for example sessions
• Run nwbinspector and fix issues
• Create demo notebook (check Figures 4 and 5 in the manuscript) Add tutorial notebook for 001084 dandi/example-notebooks#97
• Create landing page for dataset on DANDI, populate with metadata from the manuscript
• Upload to DANDI https://dandiarchive.org/dandiset/001084/draft

[weiglszonja]

For the meeting:

• They have the motion corrected image data, we need to ask if they also the x, y shifts (so that we can add this to a MotionCorrection object): https://github.com/HoweLab/MultifiberProcessing/blob/main/motion_correct.m#L356C1-L356C105 -> saving it to a .mat file https://github.com/HoweLab/MultifiberProcessing/blob/d7b33e23ed778c8c919ca90a21df0ade4b77f0e7/motion_correct.m#L356C1-L356C105

• starting time of motion corrected data should be aligned with the imaging data

• in cases where we have multiple avi files, how to automatically determine which TTL signal belongs to which avi file