Subregion specific dynamics of striatal dopamine
Data files
Abstract
Transient increases in dopamine within the striatum can encode reward prediction errors, critical signals for updating predictions of future rewards. However, it is unclear how this mechanism can provide suitable feedback for predictions across a wide range of time horizons: from seconds or less (if singing a song) to potentially hours or more (if hunting for food). Here we report that dopamine transients in distinct striatal subregions convey prediction errors over distinct time scales. Dopamine dynamics systematically accelerated from ventral to dorsal- medial to dorsal-lateral striatum, in the tempo of their spontaneous fluctuations, their temporal integration of prior rewards, and their discounting of future rewards. This spectrum of time scales for evaluative computations can help achieve efficient learning and adaptive motivation for a wide range of behaviors.
README: Subregion Specific Dynamics of Striatal Dopamine
https://doi.org/10.5061/dryad.00000008m
Description of the data and file structure
The dataset included in this repository follows the standard format used by the Berke lab. It contains the necessary files and data required to replicate the experiments and findings described in the research article. Additionally, we have provided MATLAB code that can be utilized to read and interpret the data, as well as generate figures related to our results.
The dataset consists of three *.zip files. Each zip file contains MATLAB/python scripts and data files required to reproduce the graphs in corresponding figures. Following is the description for each data file.
Fig2.zip:
This text contains data and MATLAB scripts to replicate Fig2. The data is saved in .mat format, which can be loaded in MATLAB as a 1x41 struct. Each struct has recordings from one behavioral session of the bandit task and has these elements:
- Fs: Sampling Frequency of the photometry signal
- boxts: timestamp of event onset (s)
- box: metadata array with information about the recording session
behavData: Data structure with arrays that summarize measured behavioral variables in each session, such as:
- Subject: Name of the subject
- Date: Date of the experiment
- Time: Time of the trial
- Attempt: Attempt number
- Trial: Trial number
- Food: Binary variable that reports whether a trial was rewarded
- Block: Block number
- Block trial: Trial number within the block
- Center: Number of the center poke for each trial
- Port poked: Number of the poked port
- RT: Reaction Time (ms)
- MT: Movement Time (ms)
- PreT: Pre Tone period (ms)
- ResponseLimit: Response limit time (ms)
- TrPerBlock: maximum number of trials per block
- FalseStart: Number of false starts
- WrongStart: Number of failed starts
- FailureToRespond: Number of failures to start a trial
- Prob Left: Probability of reward on the left side
- Prob Right: Probability of reward on the right side
- Target Poke: Poked port
- CNI Latency: Latency to initiate the task (ms)
- SI Duration: Side-in duration (ms)
- ITI: Intertrial Interval (ms)
- TTR: Time to Reward (ms)
- sig: Light signal
- Location: Confirmed location of the fiber photometry optical fiber
- Left: binary variable that indicates which hemisphere the data was recorded from
Fig3.zip
This file contains data, MATLAB scripts and Jupiter notebook to replicate Fig3. The data is saved in .mat format, which can be loaded in MATLAB. Each mat file includes 15 recording sessions for a rat performing the conditioning task.\
For rats IM1277, IM1278, IM1299, IM1300, IM1301, IM1358, IM1359, IM1366, IM1367, IM1381, IM1382, there are two simultaneous recordings, and each row of the corresponding mat files represents a recording session and has the following relevant elements:
Fs: Sampling Frequency
Channels: data for two signal channels and two reference channels
boxts: timestamp of event onset
box: metadata array with information about the recording session
behavData: Data structure with arrays that summarize measured behavioral variables in each session
For rats IM1413, IM1414, M1415, there are three simultaneous recordings, and each row of the corresponding mat files represents a recording session and has the following relevant elements:
ratConditioning: task parameters, similar to behavData in the 10 earlier rats
boxts: timestamp of event onset
box: metadata array with information about the recording session
DLS: signal and reference recordings at DLS, and cue onset time
DMS: signal and reference recordings at DMS, and cue onset time
VS: signal and reference recordings at VS, and cue onset time
Fig5.zip
This file contains data and MATLAB scripts to reproduce Fig5. The data is stored in .mat format, which can be imported into MATLAB as a 1x5 struct (one for each subject). Each struct contains data from three behavioral sessions of the task and has the following fields:
dF1: Response to cue onset. This is a 3D array with trials in the first dimension, time in the second dimension, and subregions in the third dimension.
dF2: Response to click. This has the same structure as dF1.
foodEntry: Port occupancy for each trial.
ind: Trial type.
rew: A binary variable that indicates whether a trial was rewarded or not.
t_axis: Time axis (in seconds) for dF responses.