Pacman Behavioral Analyses

Overview

Analysis scripts (mostly in R) for the behavior and statistical analysis of the PacMan project. While Staveland_et_al_Pacman_Neural_Analyses has the python scripts for analyses that are (nearly) only brain data, these analyses include both pure behavioral analyses and both brain/behavior models.

Repository Structure

R/

This folder holds all of the helper functions that are used throughout the analyses. The functions are sourced at the beginning of each .Rmd file.

matlab_helpers/

This folder holds the matlab helper functions that are used to convert the .mat files from our WashU site to a csv with MNI coordinates for each subject

• washu_to_mni.m: This file converts the washu coordinates to MNI coordinates.

analysis/

This folder holds all of the analyses for the paper. Each subfolder corresponds to a different analysis group.

anatomical_plotting

This folder holds the scripts for creating the MNI coordinate CSV files.

• create_mni_csv.Rmd: This script creates a CSV file with the MNI coordinates for each electrode.

attack

This folder holds the analyses related to ghost attacks. (Figure 5 analyses)

• fig5_left_vs_right_mfg.Rmd: This script creates the plots for the left vs. right MFG activity during ghost attacks.
• fig5_right_mfg_models_and_figures.Rmd: This script runs the models and creates the plots for the right MFG activity during ghost attacks.
• limbic_theta_during_attack.Rmd: This script runs the models and creates the plots for the limbic theta activity during ghost attacks.
• mfg_hfa_attack.Rmd: This script runs the models for the MFG HFA activity during ghost attacks.

behavior

This folder holds the behavioral modeling and plotting. (Figure 1 analyses)

• brms_behavioral_modeling.Rmd: This script runs the Bayesian models of the behavior.
• figure_1_behave_plots.Rmd: This script creates the plots for Figure 1.
• normative_behavior_new_sample.Rmd: This script runs the normative behavior analyses on a new sample of subjects.
• normative_behavior.Rmd: This script runs the normative behavior analyses for the pilot subjects.

cleaning_ieeg_behavior

This folder holds the scripts for cleaning and combining the behavioral data for each subject.

• cleaning_behave_SUBJECT.Rmd: These scripts clean the behavioral data for each subject.
• combine_ieeg_subs_behavior.Rmd: This script combines the cleaned behavioral data for all subjects.

clinical

This folder holds the exploratory analyses of the clinical cohort. (Supplement)

• clinical_eda.Rmd: This script performs exploratory data analysis on the clinical cohort, which is included as a supplemental figure.

coherence

This folder holds the coherence analyses. (Figure 3 analyses)

• create_sig_theta_coherence_csv.Rmd: This script creates a CSV file with the significant theta coherence values.
• coherence_region_models.Rmd: This script runs the coherence models for the different regions.
• coherence_region_models_ppc_plv.Rmd: This script runs the coherence models for the PPC and PLV.
• fig3_simplified_differential_conn_plots.Rmd: This script creates the simplified differential connectivity plots for Figure 3.
• supp_fig5_differential_conn_plots.Rmd: This script creates the differential connectivity plots for Supplementary Figure 5.
• supp_fig6_differential_conn_plots_ppc_plv.Rmd: This script creates the differential connectivity plots for the PPC and PLV for Supplementary Figure 6.
• supp_table_coherence_region.Rmd: This script creates the supplementary table for the coherence region models.
• fig_3_rising_falling_plots.Rmd: This script creates the plots for the rising and falling coherence.
• fig_3_rising_falling_regional_plots.Rmd: This script creates the regional plots for the rising and falling coherence.
• fig3_example_coherence.Rmd: This script creates the example coherence plots for Figure 3.
• fig3_rising_falling_example_sub.Rmd: This script creates the example rising and falling coherence plots for a single subject.
• rising_falling_threshold_model_comparison.Rmd: This script compares the rising and falling threshold models.
• supp_table_rising_falling.Rmd: This script creates the supplementary table for the rising and falling coherence models.

cross_correlations

This folder holds the computation and statistical analysis of cross-correlations for the directionality analysis. (Figure 4 analyses)

• compute_regional_ccfs_theta.Rmd: This script computes the regional cross-correlations for theta.
• theta_ccf_stats.Rmd: This script runs the statistics on the theta cross-correlations.

freq_power_analyses

This folder holds the analyses of theta and HFA power profiles. (Figure 2 analyses)

• compile_ieeg_files_iti_logged_trialonset.Rmd: This script compiles the iEEG files for the ITI and trial onset.
• theta_and_hfa_power_profiles.Rmd: This script creates the theta and HFA power profiles.
• all_roi_theta_app_av.Rmd: This script runs the models for the theta power in approach vs avoidance windows in all ROIs.

granger

This folder holds the Granger causality analyses. (Figure 4 analyses)

• granger_lmes.Rmd: This script runs the linear mixed effects models on the Granger causality results to determine directionality.
• granger_plots.Rmd: This script creates the plots for the directionality results.
• granger_thresholds.Rmd: This script tests the Granger causality results at different theta coherence thresholds.

turnaround_time_correlations

This folder holds the analyses of the correlation between turnaround time and theta/HFA syncrhony. (Figure 4 analyses)

• brms_all_roi_model.Rmd: This script runs the Bayesian models theta syncrhony across threshold.
• brms_hfa_all_roi_model.Rmd: This script runs the Bayesian models for HFA in all ROIs.
• calculate_trial_correlations.Rmd: This script calculates the trial-by-trial correlations.
• Figure4_combinedhfa_theta_plots.Rmd: This script creates the combined HFA and theta plots for Figure 4.
• supp_table_turn_times.Rmd: This script creates the supplementary table for the turnaround times.
• turntime_threshold_model_comparison.Rmd: This script compares the syncrhony results across different thresholds.

Run Time

For one patient with an average number of electrodes, statistical analysis, including the construction of null distributions, should be under two weeks.

Installation

Requirements

This code was created using R v4.1.2. All required packages are loaded at the beginning of each script.

Key Dependencies

• tidyverse: For data manipulation and visualization.
• brms: For Bayesian modeling.
• ggplot2: For plotting.

Install time (for repo)

Under 5 minutes, assuming main dependencies are already installed