Organization of reward and movement signals in the basal ganglia and cerebellum
Data files
Jan 31, 2024 version files 41.09 GB
-
Data.zip
-
README.md
-
task_info.mat
Abstract
The basal ganglia and the cerebellum are major subcortical structures in the motor system. The basal ganglia have been cast as the reward center of the motor system, whereas the cerebellum is thought to be involved in adjusting sensorimotor parameters. Recent findings of reward signals in the cerebellum have challenged this dichotomous view. To compare the basal ganglia and the cerebellum directly, we recorded from oculomotor regions in both structures from the same monkeys. We partitioned the trial-by-trial variability of the neurons into reward and eye-movement signals to compare the coding across structures. Reward expectation and movement signals were the most pronounced in the output structure of the basal ganglia, intermediate in the cerebellum, and the smallest in the input structure of the basal ganglia. These findings suggest that reward and movement information is sharpened through the basal ganglia, resulting in a higher signal-to-noise ratio than in the cerebellum.
README
The data is organized in two folders corresponding to the two tasks described in detail in the Methods section of the paper.
Data Structure Documentation
Each file corresponds to a single neuron recorded in a single task. The file is a structure called 'data' and contains two important fields:
1. info
2. trials
Each field is itself a structure.
info contains general information about the neuron:
1. cell_ID – an identifier for the neuron.
2. session – an identifier for the session in which the neuron was recorded
3. cell_type – The population this neuron belongs to.
4. electrode – identifier of the electrode the neuron was recorded on.
5. grade – a number that indicates the quality of sorting, assigned manually (5 – best, 10 – worst. We usually only use neurons with a grade below 7).
6. task – the task the monkey was performing.
As the name suggests, "trials" contain the trials. The fields are as following:
1. maestro_name – identifier for the trial that can be used to recognize it, when looking for trials recorded in pairs of neurons, or the behavior in the same session.
2. name – the is the type of trial. It can be used to get information about what happened in this trial.
a. The number after the "d" is the direction in which the target moved.
b. The number after the "v" is the velocity at which the target moved, if there is no v, it means that this is a saccade trial.
c. The number after "P" is the reward probability color.
d. "R" means this is a rewarded trial, "NR" means that this is an omission trial.
3. fail – whether the monkey failed this trial (boolean).
4. movement_onset – the time from the trial start in which the target began to move or jumped to an eccentric location (ms).
5. movement_offset – the time from the trial start in which the target stopped moving (ms).
6. cue_onset - the time from the trial start in which the color of the target changed (ms).
7. blinkBegin – the times from the trial start at the beginning of a blink was detected (ms).
8. blinkEnd – the times from the trial start at the end of a blink was detected (ms).
9. beginSaccade– the times from the trial start at the beginning of a saccade was detected (ms).
10. beginEnd – the times from the trial start at the end of a saccade was detected (ms).
11. spike_times – the times from the trial start at in which the neuron spikes (ms).
12. rwd_time_in_extended – times in which the neuron spiked, this is relative to a window around the trial and is separated from spike_times for technical reasons. This can be used for the outcome session.
13. extended_spike_times – times in which the neuron spiked, this is relative to a window around the trial and is separated from spike_times for technical reasons. This can be used for the outcome session (ms).
14. rwd_time_in_extended – time in which the reward was delivered or omitted, relative to the same time as extended_spike_times (ms).
15. previous_completed – whether or not the previous trial was completed successfully (boolean).
16. extended_saccade_begin - the times at which the beginning of a saccade was detected (ms), relative to the same time as extended_spike_times (ms).
17. extended_saccade_end - the times at which the end of a saccade was detected (ms), relative to the same time as extended_spike_times (ms).
18. screen_rotation – will be non-zero in case the image on the screen was rotated, this would be zero for all all trials in this dataset.
Each data file has a corresponding behavior file that can be found in the "behavior" folder. The name of the corresponding behavior file can be found in the corresponding data file (data.info.behavior_shadow_name). These files contain the eye movement velocity and position during the trials in data.trials. The behavior files are structures called 'behavior_data' and contain two fields:
1. info – same as the data files
2. trials
behavior_data.trials contain the fields:
1. maestro_name – an identifier for the trial.
2. hPos – Horizontal eye position from the trial beginning (deg/s, sampling rate: 1ms).
3. vPos – Vertical eye position.
4. hVel - Horizontal eye velocity.
5. vVel - Vertical eye velocity.
For any questions or clarifications, please reach out to noga.larry@mail.huji.ac.il
Methods
See doi: https://doi.org/10.1101/2022.10.01.510246 for complete Methods information.