This repository contains MATLAB scripts used to conduct border/ectopic variant analyses from Dworetsky et al., 2024, Two common and distinct forms of variation in human functional brain networks (Nature Neuroscience). These scripts allow users to classify network variants as either border shifts or ectopic intrusions, and compare their spatial/network distributions, prediction of behavioral phenotypes, activation during tasks, subgroup properties, and genetic similarity.

For original variant creation steps, see scripts in GitHub repo from Seitzman et al., 2019, Trait-like variants in human functional brain networks (PNAS): https://github.com/GrattonLab/SeitzmanGratton-2019-PNAS to compute individual-to-group spatial correlation, binarize pre-variant maps, and extract variants' correlation coefficients to network templates.

Once variants are created, the following script should be run first:

• variant_classification: First, classify variants as border or ectopic using classify_variants_network_dependent.m (parcellation-dependent, distance-to-same-network classification) or classify_variants_parcellation_free.m (parcellation-independent, distance-to-peak-group-similarity classification).

The scripts in remaining folders can then be run in any order:

• spatial_distribution: perform and visualize results of permutation analysis examining border/ectopic variant differences in spatial distribution across the cortex, plus cluster-correction of permutation analyses.
• network_distribution: perform and plot results of permutation analysis examining border/ectopic variant differences in network assignments.
• twins: compute and plot Dice coefficient of subject pairs of interest across groups (e.g., MZ twins, DZ twins, siblings, unrelated individuals); permutation analysis of Falconer's formula in MZ and DZ groups.
• task_responses: organize task contrast map data; compare border and ectopic task activations across contrasts in relation to canonical assigned network and all other networks; plot normalized value of variants' activations shift toward canonical network (for networks of interest).
• behavior: use variants' network or location info to predict behavioral phenotypes (see README-behprediction.txt for more info).
• subgroups: identify/plot properties of border/ectopic subgroups across individuals. Intended to be run after templateMatchingVariants.m from https://github.com/GrattonLab/SeitzmanGratton-2019-PNAS (separately for border and ectopic variants). Also requires Infomap (https://www.mapequation.org) - Run_Infomap_2015.m has been previously shared at https://github.com/MidnightScanClub/MSCcodebase/tree/master/Utilities/Infomap_wrapper.

Resources:

Each of these scripts requires supporting scripts for reading and writing CIFTI files (see more at https://github.com/fieldtrip/fieldtrip/). We have included modified versions of these scripts in the resources folder. This folder also includes a scripts for computing Dice coefficients and a script for plotting jittered distributions.