Best way of computing graph attributes by lobe

[Gershon]

Dear Chris,

I would like to compute the same graph attributes by lobe. I am using the Schaefer atlas and am coding the different networks on to the ‘lobe’ column. I might have the hypothesis that attributes of certain networks might be affected to a greater degree, rather than the whole brain. To interrogate this question, I would like to compute the same graph metrics (global efficiency, small worldness, etc…) but do so for each network/lobe). What would be the most efficient way of doing this. 

I was thinking doing so at the stage of creating the correlation matrix. Ie.,

corrs <- corr.matrix(all.dat.resids, densities=densities, exclude.reg=exclude.reg)

Using the exclude.reg function, I could then only include regions pertaining to a particular network. Is this the most efficient way to apply this type of analysis using brainGraph?

Best,

Gershon.

--
DOCTOR GERSHON SPITZ
NHMRC Early Career Fellow
BA (Hons), PhD
Monash Epworth Rehabilitation Research Centre 

Turner Institute for Brain and Mental Health
Brain Injury and Rehabilitation Theme
Brain Mapping and Modelling Theme
Monash University
Clayton campus VIC 3800
Australia 

Wellcome Centre for Integrative Neuroimaging (WIN)
FMRIB
Nuffield Department of Clinical Neurosciences
University of Oxford

@gershonspitz

[Chris Watson]

It depends on what you want to compute, exactly.

* If you are interested in a subnetwork consisting only of regions for a given lobe, then yes your proposed approach would be fine.

* If you want to be sure that inter-lobe connections are not ignored, then it would be better to construct the entire graphs and then create a data.table w/ vertex-level metrics, and calculate by lobe using data.table syntax. e.g., something like

DT <- vertex_attr_dt(g)

DT[, mean(E.nodal.wt), by=lobe]

Chris

--
You received this message because you are subscribed to the Google Groups "brainGraph-help" group.
To unsubscribe from this group and stop receiving emails from it, send an email to brainGraph-he...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/brainGraph-help/7E59ED58-3632-4602-BDCF-D8E76F8709C9%40monash.edu.

[Gershon Spitz]

Thanks Chris,

Best

<Outlook-k3nmk5tv.png>

@gershonspitz

--
You received this message because you are subscribed to the Google Groups "brainGraph-help" group.
To unsubscribe from this group and stop receiving emails from it, send an email to brainGraph-he...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/brainGraph-help/7E59ED58-3632-4602-BDCF-D8E76F8709C9%40monash.edu.

--
You received this message because you are subscribed to the Google Groups "brainGraph-help" group.
To unsubscribe from this group and stop receiving emails from it, send an email to brainGraph-he...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/brainGraph-help/CAHVz74i974bAFYpFK3%3DD9XTkAM9BvPjgGRWKhfGJhtetQQswmg%40mail.gmail.com.