1. General Description

This repository contains the processed EEG microstate sequences, analysis code (MATLAB), and figure generation data associated with the manuscript mentioned above. The study investigates the temporal dynamics of EEG microstates during different states of consciousness using Chaos Game Representation (CGR) and multimer analysis.

The data is organized into three main datasets (Datasets 1, 2, and 3).

2. Folder Structure Overview

The main compressed file Data_and_codes.zip contains the following directories:

• Dataset1/: Contains microstate sequences (.txt), analysis scripts (.m), and result variables (.mat) for the sedation dataset.
• Dataset2/: Contains data and scripts for the sedation dataset.
• Dataset3/: Contains data and scripts for the sleep dataset.
• Figures/: Contains specific data and code used to generate the figures presented in the manuscript.
• Functions/: Contains common MATLAB functions required by the analysis scripts.

3. Detailed Data Description

A. Datasets (Dataset1, Dataset2, Dataset3)

Each Dataset folder contains three types of files:

1. Microstate Sequences (.txt)
   • Description: These files contain the discrete microstate class labels derived from the preprocessed EEG data.
   • Format: Text files where each row represents an observation, and the character (e.g., A, B, C, D) represents the assigned microstate class map.
2. Analysis Scripts (.m)
   • Description: MATLAB scripts used to process the .txt sequences.
   • Function: These scripts calculate metrics such as Chaos Game Representation (CGR) spectrum, multimer duration, coverage, occurrence, and monomer count.
3. Result Variables (.mat)
   • Description: MATLAB binary files containing the computed statistics used for final statistical analysis and plotting.

B. Functions Folder (/Functions)

This folder contains necessary functions called by the main analysis scripts. Ensure this folder is added to your MATLAB path before running any analysis. Key functions include:

• chaos_game_mapping.m / chaos_game_mapping_ngon.m: Implements the CGR algorithm to map microstate sequences into 2D space.
• calc_multimer_stats.m: Identify and extract multimer patterns and their dynamic metrics from sequences using the algorithm described in the manuscript.

C. Figures Folder (/Figures)

This folder is organized by the figures as they appear in the manuscript. Each subfolder contains the specific .mat data and .m plotting script to reproduce that figure.

• Figure1/:
  • Contains code/data for Figure 1B & 1C.
  • Context: Illustrates the CGR-based spectral analysis framework and group-level spectra across frequency bands and consciousness states.
• Figure2/:
  • Contains code/data for Figure 2.
  • Context: Comparison of periodic dynamics (Peak Power and Center Frequency) in EEG microstate sequences across Datasets 1, 2, and 3.
• Figure3/:
  • Contains code/data for Figure 3A, 3B, & 3E.
  • Context: Shows temporal regularities, comparing empirical data to shuffled/Markov surrogates, and illustrates multimer coverage across Datasets 1, 2, and 3.
• Figure4/:
  • Contains code/data for Figure 4B & 4C.
  • Context: CGR spectra of multimer-constrained surrogate sequences and temporally smoothed sequences (Dataset 1).
• Figure5/:
  • Contains code/data for Figure 5B.
  • Context: Reproduction of spectral periodicity using a 3-layer hierarchical generative model.
• Figure67/:
  • Contains code/data for Figure 6 and Figure 7.
  • Context (Fig 6): Statistical comparison of multimer duration and occurrence across consciousness states.
  • Context (Fig 7): Statistical comparison of multimer coverage and monomer counts.
• FigureS1/:
  • Context: Group-level CGR spectra for Datasets 2 & 3 (Empirical vs. Shuffled vs. Markov).
• FigureS2/:
  • Context: Surrogate analysis and temporal smoothing results for Datasets 2 & 3.
• FigureS3/:
  • Context: Generative model results for Datasets 2 & 3.
• FigureS456/:
  • Context: Empirical distributions of microstate durations compared with exponential and log-normal fits for Datasets 1 (Fig S4), 2 (Fig S5), and 3 (Fig S6). Uses AIC for goodness of fit.
• FigureS7/:
  • Context: Validation of the CGR framework using six microstate classes (instead of four), focusing on Beta-band dynamics.

4. Source Data Citations

The microstate sequences in this repository were derived from the following publicly available or previously published EEG datasets. Please cite the original sources if you use the raw data:

Dataset 1 Source:

Chennu, S., O'Connor, S., Adapa, R., Menon, D.K., Bekinschtein, T.A., 2016. Brain Connectivity Dissociates Responsiveness from Drug Exposure during Propofol-Induced Transitions of Consciousness. PLoS Comput Biol 12, e1004669.

Dataset 2 Source:

Bajwa, I.J., Nilsen, A.S., Skukies, R., Aamodt, A., Ernst, G., Storm, J.F., Juel, B.E., 2025. A repeated awakening study exploring the capacity of complexity measures to capture dreaming during propofol sedation. Sci Rep 15, 32746.

Dataset 3 Source:

Wei, X., Avigdor, T., Ho, A., Minato, E., Garcia-Asensi, A., Royer, J., Wang, Y.L., Travnicek, V., Schiller, K., Bernhardt, B.C., Frauscher, B., 2024. ANPHY-Sleep: an Open Sleep Database from Healthy Adults Using High-Density Scalp Electroencephalogram. Sci Data 11, 896.

5. Software Requirements

• MATLAB: The code was developed and tested using MATLAB (MathWorks).

If you have any questions regarding the code or data, please contact: zhoudongdong@cqmu.edu.cn