WGCNA: nearestNeighborConnectivity – R documentation

Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!

nearestNeighborConnectivity

Connectivity to a constant number of nearest neighbors

Description

Given expression data and basic network parameters, the function calculates connectivity of each gene to a given number of nearest neighbors.

Usage

nearestNeighborConnectivity(datExpr, 
         nNeighbors = 50, power = 6, type = "unsigned", 
         corFnc = "cor", corOptions = "use = 'p'", 
         blockSize = 1000,
         sampleLinks = NULL, nLinks = 5000, setSeed = 38457,
         verbose = 1, indent = 0)

Arguments

`datExpr`	a data frame containing expression data, with rows corresponding to samples and columns to genes. Missing values are allowed and will be ignored.
`nNeighbors`	number of nearest neighbors to use.
`power`	soft thresholding power for network construction. Should be a number greater than 1.
`type`	a character string encoding network type. Recognized values are (unique abbreviations of) `"unsigned"`, `"signed"`, and `"signed hybrid"`.
`corFnc`	character string containing the name of the function to calculate correlation. Suggested functions include `"cor"` and `"bicor"`.
`corOptions`	further argument to the correlation function.
`blockSize`	correlation calculations will be split into square blocks of this size, to prevent running out of memory for large gene sets.
`sampleLinks`	logical: should network connections be sampled (`TRUE`) or should all connections be used systematically (`FALSE`)?
`nLinks`	number of links to be sampled. Should be set such that `nLinks * nNeighbors` be several times larger than the number of genes.
`setSeed`	seed to be used for sampling, for repeatability. If a seed already exists, it is saved before the sampling starts and restored upon exit.
`verbose`	integer controlling the level of verbosity. 0 means silent.
`indent`	integer controlling indentation of output. Each unit above 0 adds two spaces.

Details

Connectivity of gene i is the sum of adjacency strengths between gene i and other genes; in this case we take the nNeighbors nodes with the highest connection strength to gene i. The adjacency strengths are calculated by correlating the given expression data using the function supplied in corFNC and transforming them into adjacency according to the given network type and power.

Value

A vector with one component for each gene containing the nearest neighbor connectivity.

Author(s)

Peter Langfelder