NBS {brainGraph}

Network-based statistic for brain MRI data

Description

Calculates the network-based statistic (NBS), which allows for family-wise error (FWE) control over network data, introduced for brain MRI data by Zalesky et al. Requires a three-dimensional array of all subjects' connectivity matrices and a data.table of covariates, in addition to a contrast matrix or list. A null distribution of the largest connected component size is created by fitting a GLM to permuted data. For details, see GLM.

Usage

NBS(A, covars, contrasts, con.type = c("t", "f"), X = NULL,
  con.name = NULL, p.init = 0.001, perm.method = c("freedmanLane",
  "terBraak", "smith", "draperStoneman", "manly", "stillWhite"),
  part.method = c("beckmann", "guttman", "ridgway"), N = 1000,
  perms = NULL, symm.by = c("max", "min", "avg"),
  alternative = c("two.sided", "less", "greater"), long = FALSE, ...)

## S3 method for class 'NBS'
summary(object, contrast = NULL, digits = max(3L,
  getOption("digits") - 2L), ...)

## S3 method for class 'NBS'
nobs(object, ...)

## S3 method for class 'NBS'
terms(x, ...)

## S3 method for class 'NBS'
formula(x, ...)

## S3 method for class 'NBS'
labels(object, ...)

## S3 method for class 'NBS'
case.names(object, ...)

## S3 method for class 'NBS'
variable.names(object, ...)

## S3 method for class 'NBS'
df.residual(object, ...)

## S3 method for class 'NBS'
nregions(object)

Arguments

A

Three-dimensional array of all subjects' connectivity matrices

covars

A data.table of covariates

contrasts

Numeric matrix (for T statistics) or list of matrices (for F statistics) specifying the contrast(s) of interest; if only one contrast is desired, you can supply a vector (for T statistics)

con.type

Character string; either 't' or 'f' (for t or F-statistics). Default: 't'

X

Numeric matrix, if you wish to supply your own design matrix. Ignored if outcome != measure.

con.name

Character vector of the contrast name(s); if contrasts has row/list names, those will be used for reporting results

p.init

Numeric; the initial p-value threshold (default: 0.001)

perm.method

Character string indicating the permutation method. Default: 'freedmanLane'

part.method

Character string; the method of partitioning the design matrix into covariates of interest and nuisance. Default: 'beckmann'

N

Integer; number of permutations to create. Default: 5e3

perms

Matrix of permutations, if you would like to provide your own. Default: NULL

symm.by

Character string; how to create symmetric off-diagonal elements. Default: max

alternative

Character string, whether to do a two- or one-sided test. Default: 'two.sided'

long

Logical indicating whether or not to return all permutation results. Default: FALSE

...

Arguments passed to brainGraph_GLM_design

object, x

A NBS object

contrast

Integer specifying the contrast to plot/summarize; defaults to showing results for all contrasts

digits

Integer specifying the number of digits to display for P-values

Details

When printing a summary, you can include arguments to printCoefmat.

Value

An object of class NBS with some input arguments in addition to:

X

The design matrix

removed.subs

Character vector of subject ID's removed due to incomplete data (if any)

T.mat

3-d array of (symmetric) numeric matrices containing the statistics for each edge

p.mat

3-d array of (symmetric) numeric matrices containing the P-values

components

List containing data tables of the observed and permuted connected component sizes and P-values

rank, df.residual, qr, cov.unscaled

The rank, residual degrees of freedom, QR decomposition, and unscaled covariance matrix of the design matrix

Note

It is assumed that the order of the subjects in covars matches that of the input array A. You will need to ensure that this is the case. Prior to v3.0.0, the covars table was sorted by Study.ID before creating the design matrix.

Author(s)

Christopher G. Watson, [e-mail]

References

Zalesky, A. and Fornito, A. and Bullmore, E.T. (2010) Network-based statistic: identifying differences in brain networks. NeuroImage, 53(4), 1197–1207. doi: 10.1016/j.neuroimage.2010.06.041

See Also

Other Group analysis functions: Bootstrapping, GLM, Mediation, brainGraph_permute, mtpc

Examples

## Not run: 
max.comp.nbs <- NBS(A.norm.sub[[1]], covars.dti, N=5e3)

## End(Not run)