ics: Induced Conserved Structure (ICS)

In langendorfr/netcom: NETwork COMparison Inference

Induced Conserved Structure (ICS)

Description

Calculates the Induced Conserved Structure proposed by Patro and Kingsford (2012) of an alignment between two networks.

Usage

ics(network_1_input, network_2_input, alignment, flip = FALSE)

Arguments

network_1_input	The first network being aligned, which must be in matrix form. If the two networks are of different sizes, it will be easier to interpret the output if this is the smaller one.
network_2_input	The second network, which also must be a matrix.
alignment	A matrix, such as is output by the function NetCom, where the first two columns contain corresponding node IDs for the two networks that were aligned.
flip	Defaults to FALSE. Set to TRUE if the first network is larger than the second. This is necessary because ICS is not a symmetric measure of alignment quality.

Value

A number ranging between 0 and 1. If the Induced Conserved Structure is 1, the two networks are isomorphic (identical) under the given alignment.

References

Patro, R., & Kingsford, C. (2012). Global network alignment using multiscale spectral signatures. Bioinformatics, 28(23), 3105-3114.

Examples

# Note that ICS is only defined on unweighted networks.
net_one <- round(matrix(runif(25,0,1), nrow=5, ncol=5))
net_two <- round(matrix(runif(25,0,1), nrow=5, ncol=5))

ics(net_two, net_two, align(net_one, net_two)$alignment)

langendorfr/netcom documentation built on July 23, 2022, 5:19 p.m.