Research Articles

Research Articles

Data files for "Simple models of human brain functional networks"

The MATLAB workspace 'NV_COS_data.mat' contains data for the group of 20 healthy volunteers and 19 participants with Childhood Onset Schizophrenia (COS).

'data.coord' is a 140x3 array containing the spatial coordinates of the centroid of all 140 brain regions considered.

'data.EucD' is a 140x140 matrix containing the euclidian distance between the centroids of all 140 brain regions considered.

'NVdata.timeseries' is a 1x20 cell array, where each cell corresponds to one subject, and contains a 140x134 array of time series. This corresponds to time series of length 134 for each of the 140 brain regions considered. Each time series is a regional average, band-pass filtered (using a maximal overlap discrete wavelet transform with a Daubechies 4 wavelet) to the frequency interval: 0.05-0.111 Hz (scale 2).

'NVdata.matrix_corr' is a 1x20 cell array, where each cell corresponds to one subject, and contains a 140x140 correlation matrix resulting from correlating the time series from 'NVdata.timeseries'.

'NVdata.matrix_thresh_0p04' is a 1x20 cell array, where each cell corresponds to one subject, and contains a 140x140 matrix resulting from thresholding the 'NVdata.matrix_corr' matrices to cost = 0.04. As described in the methods: we used a minimum spanning tree (MST) as a backbone, then added further edges in order of decreasing correlation strength until 4% cost was reached.

Before thresholding, diagonal elements were set to zero and only the absolute value of correlations was retained.

'COSdata.matrix_thresh_0p04'is a 1x19 cell array analogous to 'NVdata.matrix_thresh_0p04', but for the 19 participants with COS.

Download: Simple_Network_Models.tar.gz file.