ABCDEFGHIJKLMNOPQRSTUVWXYZAAABACADAEAFAGAHAIAJAKALAMANAOAP
1
THIS DOCUMENT CONTAINS SUGGESTED FILENAMES FOR OPEN NEUROPHYSIOLOGY ENVIRONMENT DATASETS
2
OPEN NEUROPHYSIOLOGY ENVIRONMENT DOCUMENTATION IS HERE
3
FilenameArray dimensionDescriptionUnitnotes
4
5
Spike sorting:
6
Info on spikes
7
spikes.times.npynSpikestimes of spikess, relative to experiment start
8
spikes.depths.npynSpikesdepth of spikes
um, relative to deepest site on probe, positive means above this
9
spikes.amps.npynSpikespeak amplitudeV
10
spikes.samples.npynSpikestimes of spikes in samples
At what sample in raw ephys file does the spike occur
11
spikes.templates.npynSpikestemplate assignment of spikes by automatic algorithmint, counting from 0
12
spikes.clusters.npynSpikescluster assingment of spikes (clusters can result from manual curation of automatically-assigned templates)int, counting from 0
13
14
Info on templates (produced automatically)
15
templates.amps.npynTemplatespeak amplitude of each template produced by automatic spike sortingV
16
templates.waveforms.npy[nTemplates, nSamples, nChSub]Mean waveform of template on subset of channels V
17
templates.waveformsChannels.npy[nTemplates, nChSub]Physical channel number of channels for which waveform was specifiedint, counting from 0
18
19
Info on clusters (which might be manually curated)
20
clusters.mlapdv.npy[nClusters, 3]3d location of each cluster relative to bregma (mediolateral; anteroposterior; dorsoventral)
um: positive means right, anterior, dorsal
21
clusters.brainLocationIds_ccf_2017.npynClustersBrain location ID from 2017 Allen CCF http://download.alleninstitute.org/informatics-archive/current-release/mouse_ccf/annotation/ccf_2017/annotation_25.nrrdint, counting from 0
22
clusters.brainLocationAcronyms_ccf_2017.txtnClustersBrain location acronym from 2017 Allen CCF http://download.alleninstitute.org/informatics-archive/current-release/mouse_ccf/annotation/ccf_2017/annotation_25.nrrd
23
clusters.waveforms.npy[nClusters, nSamples, nChSub]Mean waveform of cluster on subset of channels V
24
clusters.waveformsChannels.npy[nClusters, nChSub]Physical channel number of channels for which waveform was specifiedint, counting from 0
25
clusters.depths.npynClustersDepth of cluster
um, relative to deepest site on probe, positive means above this
26
clusters.peakToTrough.npynClustersPeak-to-trough time (spike width)ms
27
clusters.amps.npynClustersMean peak amplitude of all spikes in each clusterV
28
clusters.channels.npynClusterschannel number of largest amplitude for each clusterint, counting from 0
29
30
31
Info on drift tracking (produced automatically)
32
drift.um.npy[nDriftTimes, nDriftPoints]result of drift registration. Non-rigid registration is captured by tracking the drift of multiple points along the probe
um. Cells moving downwards means negative
33
drift.times.npynDriftTimestime corresponding to each row of the drift matrixs, relative to experiment start
34
drift_depths.um.npynDriftPointsdepth corresponding to each point whose drift is tracked
um, relative to probe bottom, positive means above bottom
35
36
37
Behavior tracking
38
Eye tracking
39
eye.raw.mj2[nEyeSamples, nX, nY]Raw movie data for pupil tracking
40
eye.diameter.npy[nEyeSamples]Diameter of pupil. If two-columns, this gives diameter of left and right eye. If only one column, which eye can be found in experiment-specific documentationpixels
41
eye.centerPos.npy[nEyeSamples, 2]matrix with 2 columns giving x and y position of pupil (in pixels). If 4-columns, gives xy center position of left and right eye. If only 2 columns, which eye this refers to can be found in experiment documentationpixels
42
eye.blink.npy[nEyeSamples]Boolean array saying whether eye was blinking in each framebool
43
eye.times.npynEyeSamplesTimes of each eye tracking frames, relative to experiment start
44
45
Lick detection
46
licks.times.npy[nLicks]Times of licks as detected from DLC tongue tracess, relative to experiment start
47
48
Behavioral camera analysis and DeepLabCut output
49
camera.dlc.pqt[nframes, nPoints*3]Coordinates of DeepLabCut (DLC)-detected points (x position, y position, likelihood). pixels
Columns named in the parquet file. Current IBL data has 11 points = 11 (fpaws-2, nose-1, spout-2, tongue-2, eye-4).
50
camera.times.npy[nframes]times of acquisition of all frames s, relative to experiment start
51
camera.features.pqt[nframes, nfeatures]Contains features calculated from DLC traces.
Columns named in the parquet file. Current IBL data has pupilDiameter_raw, pupilDiameter_smooth
52
camera.ROIMotionEnergy.npy[nframes, nRois]Motion energy calculated within specified ROIs
53
ROIMotionEnergy.position.npy[nRois, 4](w, h, x, y) where w and h are the width and height of the ROI, x and y are its upper left cornerpixels
54
55
Silicon probe geometry and track reconstruction
56
Info on probes
57
probes.trajectory.json[nProbes, 7]JSON with one entry per probe containing 7 parameters describing the probe trajectory.
'x':(um) medio-lateral coordinate relative to Bregma, left negative
'y':(um) antero-posterior coordinate relative to Bregma, back negative
'z':(um) dorso-ventral coordinate relative to Bregma, ventral negative
'phi':(degrees)[-180 180] azimuth
'theta':(degrees)[0 180] polar angle
'depth':(um) insertion depth
'beta' :(degrees) roll angle of the probe		um or degrees
58
probes.description.json[nProbes]JSON with one entry per probe containing label, model (3A, 3B1, 3B2),serial and raw_file_name (i.e. path to raw data file on the acquisition computer)text
59
60
61
channels.electrodeSites.npy[nch]Array of integers saying which index in the raw recording file (of its home probe) that the channel corresponds to (counting from zero). NOTE there may be less channels than electrode sites, for example reference channels may be dropped from the channels objectint counting from 0
In current IBL data this is called channels.rawInd.npy. It will be renamed in the future to channels.electrodeSites.npy
62
channels.brainLocationIds_ccf_2017.tsv[nch]Brain location id of channels following ephys alignment obtained from 25um resolution 2017 Allen Common Coordinate Framework
63
channels.mlapdv.npy[nch, 3]3d location of the channels relative to bregma following ephys alignment - mediolateral; anterior-posterior; dorsoventral coordinates (um)um relative to bregma
64
channels.localCoordinates.npy[nch, 2]Location of each channel relative to probe coordinate system (µm): x (first) dimension is on the width of the shank; (y) is the depth where 0 is the deepest site, and positive above this.um relative to probe tip
65
66
electrodeSites.localCoordinates.npy[nch, 2]Location of each channel relative to probe coordinate system (µm): (first) dimension is on the width of the shank; Second is the depth where 0 is the deepest site, and positive above this. Straight mapping to the raw electrophysiology binary files
67
electrodeSites.mlapdv.npy[nch, 3]3d location of the channels relative to bregma following ephys alignment - mediolateral; anterior-posterior; dorsoventral coordinates (um) straight mapping to the raw electrophysiology binary file channels
68
electrodeSites.brainLocationIds_ccf_2017.npy[nch, 2]Brain location id of channels following ephys alignment obtained from 25um resolution 2017 Allen Common Coordinate Framework - straight mapping to the raw electrophysiology binary file channels
69
70
71
72
73
74
75
76
77
PROPOSED FILENAMES FOR FUTURE VERSIONS
78
79
Multi-photon calcium imaging
80
Activity on each frame
81
mpci.componentsActivity.npy[nFrames, nComponents, nPlanes]SVT of singular value compression, separately for each plane. If SVT done in 3d, the nPlanes dimension is absent
For each frame f and plane p, multiplying rawActivityComponents[f,p,:] by basisImages[p,:,y,x] givens the pixel intensity of frame f, plane p, pixel (y,x), in photodetector output units
82
mpci.ROIActivityF.npy[nFrames, nROIs]mean activity of all pixels in each ROIphotodetector output units
83
mpci.ROINeuropilActivityF.npy[nFrames, nROIs]mean activity neuropil pixels neighboring each ROIphotodetector output units
84
mpci.ROIActivityDeconvolved.npy[nFrames, nROIs]neuropil-subtracted deconvolved activity of each ROI using standard parameters
AU (different deconvolution methods may give different answers)
85
mpci.times.npy[nFrames]times of each frame. Add mpcistack.offset_times to get scan time of each voxel.s, relative to experiment start
86
mpci.badFrames.npy[nFrames]bad frames as identified by processing softwareboolean
87
mpci.mpciFrameQC.npy[nFrames]frame-level quality control as added by experimenter. 0 means good, other values defined in mpciFrameQC.names.tsvint
88
mpciFrameQC.names.tsv[nQCtypes, 2]
human-readable definition of QC types. First column, 'qc_values', contains unsigned ints where 0 always should mean good. Second column, 'qc_labels', contains a short human-readable description.
string
89
mpciComponents.images.npy[nComponents, H, W, nPlanes]component image for each plane
Images normalized to have sum^2 activity of 1.
90
Summary images
91
mpciMeanImage.images.npy[nComponents, H, W, nPlanes, nChannels]mean image for each plane and each channel. Channel 0 should be the calcium-sensitive channel
92
93
Information about detected ROIs
94
mpciROIs.stackPos.npy[nROIs, 3]X, Y, and Z (plane) coordinate of each ROI's centroidpixels/planes
95
mpciROIs.mlapdv.npy[nROIs, 3]Allen CCF coordinates of each ROI centroidum, relative to bregma
96
mpciROIs.mpciROITypes.npy[nROIs]Numerical code of ROI type for each ROIint
97
mpciROIs.masks.npz[nROIs, H, W, ?nPlanes?]floating-point mask of each ROI, in 2d or 3d according to how you did detection. Saved as a sparse npz array (scipy.sparse.save_npz)float
98
mpciROIs.neuropilMasks.npz[nROIs, H, W, ?nPlanes?]floating-point neuropil mask for each ROI, in 2d or 3d according to how you did detection. Saved as a sparse npz arrayfloat
99
mpciROITypes.names.tsv[nROItypes]string describing each ROI type (neuron, dendrite, etc)string
100
mpciROIs.cellClassifier.npy[nROIS]floating-point cell classifier score for each ROI, ranging between 0 and 1float