Warning: there is a lot of unnecessary function from the legacy versions and external toolboxes which might have compatability issues/bugs. They will be deleted eventually so maybe only download the function that is relevant instead of everything.
This package is currently under development and testing. The initial commit is done by Dr. Adam Eggebrecht and is an upgrade from the [GraphTools] (https://www.nitrc.org/projects/graphtools/), with an attempt to make community detection of brain networks more principled and data-driven, especially in optimizing for the scale (number of communities).
It takes a sparsely thresholded Nroi x Nroi structural/functional connectivity matrix and find the community structure using infomap community detection (Rosvall & Bergstrom 2008 PNAS). Then it calculates the stability of solutions across algorithm initation, and across different resolution scale (by varying the threshold to the graph).
It is based on multiple lines of evidence which suggest a hierarchical organization in the brain while the original applications (e.g. Power 2011 Neuron) aimed to find a consensus across scales. I plan to determine the final communities in brain structural/functional connectivity data based on the assumption that at an optimal scale, 1) a given community detection algorithm is able to produce similar solution with each random initation, 2) the solution is similar across subsamples/bootstrapped samples of the input data, 3) the solution is similar across different subsamples/bootstraped samples of the input data, 4) clustering measures, such as silhouette index is relatively high, 5) the neighboring scales produce similar solutions and the transition is sharp across different hierarchies, 6) convergent evidence can be found from different community detection algorithm
The main goal for this repository is to explore the evidences above to find the optimal scales of human brain network organization and improves the visualization and user interaction/flexibility to run the infomap algorithm as well as alternative community detection algorithms including modularity maximization, spectral clustering, k-means clustering, weighted stochastic block model. In the future, I also intend to add detection of communities using multilayer networks to examine the linked community evolution across lifespan and/or states.
Additional goals include to correct for some bugs in the program and introduce the MST-based (Hagmann et al. 2008 PLOS Biology) and local density based (Gordon et al. 2020 PNAS) threshold such that the graph is connected at the sparsest density level and produce less noisy solutions at those densities to enable the discovery of sub-networks beyond the commonly examined scale (e.g. in Gordon et al.,2020 PNAS, sub-networks within the default mode network were delineated). And to minimize the effect of user-defined parameters on the results by examining the effect of changing those parameters and provide suggestions of the default parameters, modularize the package to separate the community detection, evaluation and selection, and visualization. Alternatively introduce Random-Matrix-Based filtering for correlation null model (MacMahon & Garlaschelli 2015 Physical Review) and fitting spatioautocorrelation for distance-based thresholding instead the current hard cut-off at 20 mm (Shinn et al. 2023 Nat Neurosci,Esfahlani et al. 2020 Neuroimage)
Also exploring different consensus including the Hierarchical Consensus idea from Jeub et al. 2018.
Finally a general purpose code to understand the solutions (e.g. different repeats or different resolution scales) is provided [currently complete part]
post_comm_assignment_ordinal.mlx Interactive tutorial for post-community detection for different resolution scales or different temporal points post_comm_assignment_categorical.mlx Interactive tutorial for post-community detection for different repeats/bootstraps/subjects/methods
- Code from other toolbox (External Functions):
- infomap (version 0.15.7 and 0.18.9))
- cifti-matlab-master
- BCT (2019_03_03 version)
- Faskowitz2018wsbmLifeSpan
- NetworkCommunityToolbox
- communityalg (N.B.: downloaded some depenendencies from here: https://github.com/CarloNicolini/algonet/tree/master)
- HierarchicalConsensus
- hierarchical-brain-networks
- GenLouvain
- multilouvainorigional C++ code from Vincent Traag
- Currently support parcel-based/ROI-based input (Nroi x Nroi).
- Full support for vertex-based and voxel-based input in progress.