Skip to main content
Dryad

Data from: Neural interactions in the human frontal cortex dissociate reward and punishment learning

Cite this dataset

Combrisson, Etienne et al. (2024). Data from: Neural interactions in the human frontal cortex dissociate reward and punishment learning [Dataset]. Dryad. https://doi.org/10.5061/dryad.jdfn2z3k4

Abstract

How human prefrontal and insular regions interact while maximizing rewards and minimizing punishments is unknown. Capitalizing on human intracranial recordings, we demonstrate that the functional specificity toward reward or punishment learning is better disentangled by interactions compared to local representations. Prefrontal and insular cortices display non-selective neural populations to reward and punishment. The non-selective responses, however, give rise to context-specific interareal interactions. We identify a reward subsystem with redundant interactions between the orbitofrontal and ventromedial prefrontal cortices, with a driving role of the latter. In addition, we find a punishment subsystem with redundant interactions between the insular and dorsolateral cortices, with a driving role of the insula. Finally, switching between reward and punishment learning is mediated by synergistic interactions between the two subsystems. These results provide a unifying explanation of distributed cortical representations and interactions supporting reward and punishment learning.

README: Data from: Neural interactions in the human frontal cortex dissociate reward and punishment learning

https://doi.org/10.5061/dryad.jdfn2z3k4

Preprocessed data containing intracranial EEG high-gamma activity during a probabilistic learning task

Data organisation

The data are organized into three folders :

- power : netcdf files of gamma power of individual subjects. Each matrix has a shape of (n_{trials}, n_{roi}, n_{times}). When loading this file using Xarray, each dimension of the matrix contains the labels (i.e. whether it's a rewarding or punishing trial, the channel name and time points).

- behavior : behavioral xlsx of length (n_{trials},). This table contains the values of the prediction error during the rewarding and punishing conditions.

- anatomy : xlsx file of length (n_{roi},) containing anatomical information (i.e. the name of the brain region) of each recording contact.

To load the gamma power using Python we recommend using the Xarray python :

```python

import xarray as xr

data = xr.load_dataarray(".../power_subject-X.nc")

Code/Software

The Python scripts and notebooks to reproduce the results presented here are hosted on Github

(https://github.com/brainets/papercode/tree/main/combrisson_pblt_2024).

Funding

Agence Nationale de la Recherche