pcit: Apply the PCIT algorithm

Description Usage Arguments Value Author(s) References Examples

Description

Given a correlation matrix the PCIT algorithm (Reverter & Chan 2008) is applied to identify significant correlations. If a parallel environment running Rmpi is detected, a parallel implementation will be run unless force.serial=TRUE

Usage

1
2
3
4
	pcit(m, force.serial=FALSE, force.parallel=FALSE, nslaves=NULL,
		verbose=getOption("verbose"),
		tol.type=c("mean", "min", "max", "median"),
		pass.type=c("file", "memory", "db"))

Arguments

m

- A correlation matrix.

force.serial

- A boolean to indicate if the serial implementation of PCIT should be forced.

force.parallel

- A boolean to indicate if the parallel implementation of PCIT should be forced.

nslaves

- The number of slaves to spawn. By default, as many slaves as possible are spawned. UNTESTED OPTION.

verbose

- A boolean to indicate if verbose output should be used.

tol.type

- The type of tolerance measure to be used in PCIT. Current options are "mean", "min" and "max".

pass.type

- The type of approach used to pass the correlation matrix from the master CPU to the slave CPUs. Current options are "file", "memory" and "db".

Value

Linear indices are returned for those correlations found to be significant.

Author(s)

Nathan S. Watson-Haigh

References

Reverter, A. & Chan, E.K., 2008. Combining partial correlation and an information theory approach to the reversed-engineering of gene co-expression networks. Bioinformatics, btn482.

Examples

1
2
3
4
	data(PCIT)
	m <- m[1:200,1:200]        # just use a small subset of the data
	
	result <- pcit(m)

Example output



PCIT documentation built on May 1, 2019, 8:10 p.m.