netneurolab/liu_fc-pyspi

Benchmarking methods for mapping functional connectivity in the brain

This repository contains code and data in support of "Benchmarking methods for mapping functional connectivity in the brain", now up on bioRxiv.

code

The code folder contains relevant scripts used to run the analyses and generate the figures.

A description of each file follows (in an order that complements the manuscript):

Scripts in italic are not necessary for running the analysis scripts as some key variables are provided in data/derivatives folder.

• 00_run_pyspi_hcp.py and 00_submit_job_hcp.sh are sample scripts that run the pyspi pipeline on HCP data. config-1-minimized.yaml is the configuration file for the pyspi pipeline.
• 00_sample_generate_data.py is a sample script that generates clean set of data (used in the main analyses) from raw pyspi outputs.
• 01_similarity.py replicates Figure 1.
• 02_network_properties.py replicates Figure 2.
• 03_many_networks.py replicates Figure 3.
• 04_individual_differences.py replicates Figure 4.
  • 04_sample_generate_fingerprinting.py is a sample script that generates fingerprinting data from raw pyspi outputs.
  • 04_sample_generate_prediction.py is a sample script that generates brain-behavior prediction data from raw pyspi outputs.
• 05_information_theory.py replicates Figure 5.
  • 05_sample_generate_phiid.m is a sample script that generates integrated information decomposition results from HCP data. The original script was taken from here, however, we recommend using the updated versions recommended in (Luppi et al., 2024).
  • 05_sample_generate_stats.py is a sample script that generates dominance analysis results.
• 06_sensitivity.py replicates Figure 6.
• utils.py contains utility functions used in the analyses.

data

The data folder contains data files used for the analyses. Due to the size limit, the raw data files are not included in this repository. You can download them from the OSF repository here.

• data/derivatives contains some key variables used in the analyses. You probably only need these to run the analysis scripts in the code folder.
• data/raw contains raw pyspi outputs and some additional data files.