development_EI_decorrelation

The repo is still a work-in-progress. It will be finalized before the publication of the paper

Developmental increase of inhibition drives decorrelation of neural activity

Mattia Chini, Thomas Pfeffer, Ileana L. Hanganu-Opatz

This repository contains the original mouse ephys data that was collected for the article "Developmental increase of inhibition drives decorrelation of neural activity". Links to other data repositories that were used in the project are available in the manuscript itself.
We provide in vivo recordings (LFP and SUA data) from the prefrontal cortex of >100 unanesthetized mice of 2 to 10 days of age.
For the subset of these mice that underwent optogenetic stimulation of interneuron activity, we also provide the respective SUA and LFP data during optogenetic manipulations.

Meta data:

• Meta data is provided directly with the respective LFP and SUA files. Both files are MATLAB structures in whose fields the meta data is provided.

Electrode channels:

• All recordings contain 16 channels recorded with a Neuronexus 4x4 4-shank electrode. Neighboring channels on the same shank are 100 micrometers apart. Channels at the same depth and on neighboring shanks are 125 micrometers apart.
• Channels 1 to 4 are the most medial ones and channels 13 to 16 the most lateral.
• The probe spans from layer II to layer V/VI of the prelimbic section of the prefrontal cortex.
• Channels 1-5-9-13 are the most superficial (closer to brain surface), and 4-8-12-16 the deepest.

Baseline:

This folder contains LFP and SUA data for undisturbed baseline activity.

LFP:

• matlab structure
• The LFP is a .mat file, with a sampling frequency of 100 Hz.
• Signal has been common average rereferenced (the median of all channels has been subtracted).
• Most recordings are ~ 1h 30' of uniterrupted recording. All recordings are at least 50' long.
• The LFP signal is in a two-dimensional matrix form. The first dimension corresponds to channels (see above), the second dimension to time.
• Meta data:
  • age: age of the mouse at which data was acquired (postnatal days)
  • fs: sampling frequency (Hz)
  • downsampling_factor: downsampling factor from acquisition sampling rate (32 kHz)
  • filter: band-pass filter applied to LFP signal

SUA:

• matlab structure
• spike_matrix: two-dimensional sparse binary matrix. The first dimension corresponds to individual single units, the second dimension corresponds to time (1ms bins, i.e. 1000 Hz sampling rate).
• channel: channels at which single units were recorded (see above for explanation).
• age: same as for LFP

Opto:

This folder contains LFP and SUA data for 3s-long ramp-like (see manuscript for details) optogenetic manipulation of interneuronal activity.

LFP:

• matlab structure, with the second dimension corresponding to the number of different stimulation sessions
• The LFP is a .mat file, with a sampling frequency of 200 Hz.
• The LFP signal is in a three-dimensional matrix form:
  • the first dimension corresponds to channels (see above)
  • the second dimension corresponds to the number of ramps (generally 30 or 60)
  • the third dimension corresponds to time (1800 data points at 200 Hz -> 9s):
    • the first 3s correspond to pre-stimulus
    • the central 3s correspond to the ramp stimulation
    • the last 3s correspond to post-stimulus
• Meta data:
  • age: age of the mouse at which data was acquired (postnatal days)
  • fs: sampling frequency (Hz)
  • exp_type: whether interneurons were stimulated or inhibited (see manuscript)
  • stim_location: whether the optogenetic stimulation was carried out on the most medial or lateral shank (see above)

SUA:

• spike_matrix_stim: two-dimensional binary matrix. The first dimension corresponds to individual single units, the second dimension corresponds to time (9s * the number of stimulations). In this variable, only spikes that occurred during the ramp stimulation (see above) have been kept.
• spike_matrix_pre: two-dimensional binary matrix. The first dimension corresponds to individual single units, the second dimension corresponds to time (9s * the number of stimulations). In this variable, only spikes that occurred during the pre-stimulus (see above) have been kept.
• channels: channels at which single units were recorded (see above for explanation). The order in which single units appear in the two spike matrices are the same.
• age: same as for LFP
• exp_type: same as for LFP

For every question and/or further data request, please don't hesitate in contacting me.