neuralib.atlas.brainrender
BrainRender Wrapper
- author:
Yu-Ting Wei
This module provides a CLI-based wrapper for brainrender. Once installed, the CLI can be invoked directly from the command line.
To see available options, use the -h flag:
neuralib_brainrender {area, roi, probe} -h
Region Reconstruction
Example: Reconstructing the Visual Cortex from specified brain regions:
neuralib_brainrender area -R VISal,VISam,VISl,VISli,VISp,VISpl,VISpm,VISpor --camera top
ROI Reconstruction
By default, coordinates are interpreted in the CCF coordinate space.
You can specify the coordinate space using the --coord-space option: {ccf, brainrender}.
NumPy File Input
Input shape: Array[float, (N, 3)], with AP, DV, and ML coordinates.
Example:
[[-3.03, 4.34, -4.50],
[-3.03, 4.42, -4.37],
...
[-2.91, 4.12, 4.85]]
Run:
neuralib_brainrender roi --file <NUMPY_FILE>
CSV File Input
Required columns: AP_location, DV_location, ML_location
┌─────────────┬─────────────┬─────────────┐
│ AP_location │ DV_location │ ML_location │
│------------ │-------------│-------------│
│ -3.03 │ 4.34 │ -4.50 │
│ -3.03 │ 4.92 │ -4.31 │
│ ... │ ... │ ... │
│ -2.91 │ 4.12 │ 4.85 │
└─────────────┴─────────────┴─────────────┘
Example:
neuralib_brainrender roi --file <CSV_FILE>
Flexible Reconstruction (Processed CSV)
Be able to reconstruct rois in a specific regions/subregions
Example of using parsed allenccf csv output
┌───────────────────────────────────┬─────────┬─────────────┬─────────────┬─────────────┬─────────┬─────────┬────────┬────────────┐
│ name ┆ acronym ┆ AP_location ┆ DV_location ┆ ML_location ┆ avIndex ┆ channel ┆ source ┆ ... │
│ --- ┆ --- ┆ --- ┆ --- ┆ --- ┆ --- ┆ --- ┆ --- ┆ --- │
│ str ┆ str ┆ f64 ┆ f64 ┆ f64 ┆ i64 ┆ str ┆ str ┆ ... │
╞═══════════════════════════════════╪═════════╪═════════════╪═════════════╪═════════════╪═════════╪═════════╪════════╪════════════╡
│ Ectorhinal area/Layer 5 ┆ ECT5 ┆ -3.03 ┆ 4.34 ┆ -4.5 ┆ 377 ┆ gfp ┆ VIS ┆ ... │
│ Perirhinal area layer 6a ┆ PERI6a ┆ -3.03 ┆ 4.42 ┆ -4.37 ┆ 372 ┆ gfp ┆ VIS ┆ ... │
│ … ┆ … ┆ … ┆ … ┆ … ┆ … ┆ … ┆ … ┆ … │
│ Ventral auditory area layer 6a ┆ AUDv6a ┆ -2.91 ┆ 3.52 ┆ 4.46 ┆ 156 ┆ rfp ┆ CA1 ┆ ... │
│ Ectorhinal area/Layer 6a ┆ ECT6a ┆ -2.91 ┆ 4.14 ┆ 4.47 ┆ 378 ┆ rfp ┆ CA1 ┆ ... │
│ Temporal association areas layer… ┆ TEa5 ┆ -2.91 ┆ 4.02 ┆ 4.55 ┆ 365 ┆ rfp ┆ CA1 ┆ ... │
└───────────────────────────────────┴─────────┴─────────────┴─────────────┴─────────────┴─────────┴─────────┴────────┴────────────┘
import polars as pl
from neuralib.atlas.ccf.classifier import RoiClassifierDataFrame
df = pl.DataFrame({
"acronym": ["RSPd", "RSPd", "VISp", "VISp"],
"AP_location": [1.2, 1.3, -2.4, -2.6],
"DV_location": [1.0, 1.1, 2.0, 2.1],
"ML_location": [0.4, -0.3, 0.2, -0.2],
"channel": ["gfp", "gfp", "rfp", "rfp"],
"source": ["CA1", "CA1", "CA3", "CA3"]
})
df = RoiClassifierDataFrame(df).post_processing().dataframe()
df.write_csv(CSV_FILE)
See also
Example (reconstruct ROI in the parahippocampal areas):
neuralib_brainrender roi --classifier-file <CSV_FILE> --region APr,ENT,HATA,PAR,POST,PRE,ProS,SUB --roi-region RHP --region-alpha 0.2 --roi-radius 20 --no-root -H right
Probe Reconstruction
Reconstruct probes (or shanks) based on trajectory labeling (e.g., DiI, DiO, or lesion tracks).
Default coordinate space: CCF
Set coordinate space using:
--coord-space {ccf, brainrender}Each shank must have 2 points: dorsal and ventral
NumPy File Input
Single shank: Array[float, (2, 3)] (dorsal and ventral 3D AP/ML/DV coordinates)
[[-3.82, 1.92, -3.12],
[-3.93, 4.36, -3.30]]
Multi-shank: Array[float, (S, 2, 3)]
[[[...], [...]],
[[...], [...]],
...]
CSV File Input
Required fields: AP_location, DV_location, ML_location, point, probe_idx
If loss either point, probe_idx field, then auto infer based on the given insertion --plane
┌─────────────┬─────────────┬─────────────┬─────────┬───────────┐
│ AP_location │ DV_location │ ML_location │ point │ probe_idx │
├─────────────┼─────────────┼─────────────┼─────────┼───────────┤
│ -3.81 │ 1.92 │ -3.12 │ dorsal │ 1 │
│ -3.93 │ 4.36 │ -3.30 │ ventral │ 1 │
│ ... │ ... │ ... │ ... │ ... │
└─────────────┴─────────────┴─────────────┴─────────┴───────────┘
Additional Options
--depth DEPTH: Depth (in µm) of the implantation from the brain surface--dye: Only reconstruct dye-labeled tracks (default includes both dye and theoretical)--remove-outside-brain: Exclude any segments outside the brain
Example: Reconstructing a 4-shank NeuroPixel probe targeting the left entorhinal cortex
neuralib_brainrender probe -F <FILE> --depth 3000 -P sagittal -R ENT -H left
Red = dye-labeled track
Black = theoretical track
Help
To explore available options for each subcommand, use the -h flag:
neuralib_brainrender probe -h