Data and code for: "Action information is integrated into entorhinal representations of conceptual space and is reflected in eye movements"

Data and code for the paper entitled:Action information is integrated into entorhinal representations of conceptual space and is reflected in eye movements Code is split into separate directories to analyse behavioural, fMRI and eyetracking data. Files are named sequentially, allowing a straightforward replication by following the ordered files. File paths are relative to the original user, but can be easily adapted by changing the path variables defined at the start of each script. All code assumes adherence to the BIDS format for fMRI data; behavioural and eyetracking data is saved separately in a pseudo-BIDS format. For details see code comments. Code was mostly run on a compute server (SLURM). Code necessary for this is labelled with a letter and ‘submit_slurm’, but can easily be adapted for use on any personal compute server or run as individual files using the ‘sys.argv’ commands from terminal. data Data is stored in /data/, divided into neural, deepMReye, behavioural and eyetracking data. All data are provided as .zip files. Models used for RSA analyses are also provided as a separate .zip file. code Code to reproduce all figures and findings can be found in the linked repository, and additionally in a zipped file here, divided into: behavioural analyses eyetracking analyses neural data analyses (including deepMReye) additional functions (e.g. reliability-based voxel selection) results Results for the RSA analyses can be found in /results/. Behavioural and eyetracking analyses can be found in Jupyter notebooks in the relevant code sections. A brief overview of scripts provided can be found below, to aid navigation. For any further information, feel free to contact Alex (alex.eperon@gmail.com). fMRI analysis Convert raw files to BIDS format and move to working directories Preprocess using fmriprep Create events files and ROIs for future analyses; define event files based on planned analysis (RSA) Run first-level GLM for 4 main conditions using nilearn Create a subject-specific version of the Juelich atlas mamimum probability maps to use for ROI analysis Segment the Juelich atlas into ROIs Create neural RDMs Compare model RDMs in entorhinal ROIs, excluding the effects of other models using partial correlation Run searchlight analyses Convert searchlight maps to MNI space Run cluster correction to check for significant clusters in searchlight maps Code to predict eye position using deepMReye toolbox (Frey, Nau and Doeller, 2021) Code to extract events from deepMReye-predicted gaze positions for use in further analyses eye analysis Data preprocessing; blink removal Create an events file to categorise data by condition Visualise data and create a big dataframe organised by subject and condition Test if eye movements are skewed right or left in x and y Test if deepMReye-predicted eye movements are skewed left or right in x Test if deepMReye-predicted eye movements are skewed left or right in y