Format

Send to

Choose Destination
Brain Imaging Behav. 2018 Aug 7. doi: 10.1007/s11682-018-9937-6. [Epub ahead of print]

Mapping correlations of psychological and structural connectome properties of the dataset of the human connectome project with the maximum spanning tree method.

Author information

1
PIT Bioinformatics Group, Eötvös University, H-1117, Budapest, Hungary. szalkai@pitgroup.org.
2
PIT Bioinformatics Group, Eötvös University, H-1117, Budapest, Hungary.
3
PIT Bioinformatics Group, Eötvös University, H-1117, Budapest, Hungary. grolmusz@pitgroup.org.
4
Uratim Ltd., H-1118, Budapest, Hungary. grolmusz@pitgroup.org.

Abstract

Genome-wide association studies (GWAS) opened new horizons in genomics and medicine by discovering novel genetic factors in numerous health conditions. The analogous analysis of the correlations of large quantities of psychological and brain imaging measures may yield similarly striking results in the brain science. Smith et al. (Nat Neurosci. 18(11): 1565-1567, 2015) presented a study of the associations between MRI-detected resting-state functional connectomes and behavioral data, based on the Human Connectome Project's (HCP) data release. Here we analyze the pairwise correlations between 717 psychological-, anatomical- and structural connectome-properties, based also on the Human Connectome Project's 500-subject dataset. For the connectome properties, we have focused on the structural (or anatomical) connectomes, instead of the functional connectomes. For the structural connectome analysis we have computed and publicly deposited structural braingraphs at the site http://braingraph.org . Numerous non-trivial and hard-to-compute graph-theoretical parameters (like minimum bisection width, minimum vertex cover, eigenvalue gap, maximum matching number, maximum fractional matching number) were computed for braingraphs of each subject, gained from the left- and right hemispheres and the whole brain. The correlations of these parameters, as well as other anatomical and behavioral measures were detected and analyzed. For discovering and visualizing the most interesting correlations in the 717 x 717 matrix, we have applied the maximum spanning tree method. Apart from numerous natural correlations, which describe parameters computable or approximable from one another, we have found several significant, novel correlations in the dataset, e.g., between the score of the NIH Toolbox 9-hole Pegboard Dexterity Test and the maximum weight graph theoretical matching in the left hemisphere. We also have found correlations described very recently and independently from the HCP-dataset: e.g., between gambling behavior and the number of the connections leaving the insula: these already known findings independently validate the power of our method.

KEYWORDS:

Braingraph; Connectome; Maximum spanning tree

PMID:
30088220
DOI:
10.1007/s11682-018-9937-6

Supplemental Content

Full text links

Icon for Springer
Loading ...
Support Center