Description

This is a Python package for estimating the location of cells recorded with Neuropixels within the mouse brain

Installation

Clone the repository:
git clone https://github.com/felsenlab/unit-localizer.git

Navigate to the root directory of the project:
cd ./unit-localizer

Execute the setup script with pip:
pip install .

Basic usage

Downloading the Allen Mouse Common Coordinate Framework (CCF)

from unit_localizer import downloadCCFDataFromOSF
downloadCCFDataFromOSF()

The function downloadCCFDataFromOSF downloads the CCF to your local machine. Alternatively, you can download the CCF from a different source using the downloadCCFDataFromAllenInstitute function. It should be the same data.

Estimating unit location with the Neuropixels trajectory explorer

from unit_localizer import localizeUnitsWithTrajectoryExplorerFile
kilosortOutputFolder = 'path/to/kilosort/output'
trajectoryExplorerFile = 'path/to/trajectory/explorere/file'
labels, points, transformed = localizeUnitsWithTrajectoryExplorerFile(
    kilosortOutputFolder=kilosortOutputFolder
    trajectoryExplorerFile=trajectoryExplorerFile
)

labels is an array of the brain structures associated with each unit. points is an array (N units x 3) with the coordinates of each unit in voxels of the Allen Common Coordinate Framework. transformed is an array with the same shape as points but indicates the location of each unit in stereotaxic coordinates in reference to bregma.

Estimating unit location with sterotaxic coordinates

from unit_localizer import localizeUnitsWithInsertionParameters
kilosortOutputFolder = 'path/to/kilosort/output'
insertionPoint = np.array([-3.9, 2.5, 0.4]) # Insertion point in sterotaxic coordinates (AP, ML, DV) in mm
insertionDepth = 3.6 # Depth of insertion along the trajectory of the insertion (in mm)
insertionAngle = 6 # Angle of insertion (in degrees)
skullThickness = 0.3 # Assumed thickness of the skull (in mm)
labels, points, transformed = localizeUnitsWithInsertionParameters(
    kilosortOutputFolder=kilosortOutputFolder,
    insertionPoint=insertionPoint,
    insertionDepth=insertionDepth,
    insertionAngle=insertionAngle,
    skullThickness=skullThickness
)

Utility functions

The estimateSpikeDepths function computes the distance of each spike from the tip of the Neurpixels probe. Please note that this function assumes all active channels are clustered at the tip of the electrode.

from unit_localizer import estimateSpikeDepths
kilosortOutputFolder = 'path/to/kilosort/output'
templateDepths, spikeDepths = estimateSpikeDepths(
    kilosortOutputFolder=kilosortOutputFolder
)