Given expression data or a similarity, the function constructs the adjacency matrix and for each node calculates its connectivity, that is the sum of the adjacency to the other nodes.
softConnectivity(
datExpr,
corFnc = "cor", corOptions = "use = 'p'",
type = "unsigned",
power = if (type == "signed") 15 else 6,
blockSize = 1500,
minNSamples = NULL,
verbose = 2, indent = 0)softConnectivity.fromSimilarity(
similarity,
type = "unsigned",
power = if (type == "signed") 15 else 6,
blockSize = 1500,
verbose = 2, indent = 0)
a data frame containing the expression data, with rows corresponding to samples and columns to genes.
a similarity matrix: a square symmetric matrix with entries between -1 and 1.
character string giving the correlation function to be used for the adjacency
calculation. Recommended choices are "cor" and "bicor", but other functions can be used as
well.
character string giving further options to be passed to the correlation function.
network type. Allowed values are (unique abbreviations of) "unsigned",
"signed", "signed hybrid".
soft thresholding power.
block size in which adjacency is to be calculated. Too low (say below 100) may make the calculation inefficient, while too high may cause R to run out of physical memory and slow down the computer. Should be chosen such that an array of doubles of size (number of genes) * (block size) fits into available physical memory.
minimum number of samples available for the calculation of adjacency for the
adjacency to be considered valid. If not given, defaults to the greater of ..minNSamples
(currently 4) and number of samples divided by 3. If the number of samples falls below this threshold,
the connectivity of the corresponding gene will be returned as NA.
integer level of verbosity. Zero means silent, higher values make the output progressively more and more verbose.
indentation for diagnostic messages. Zero means no indentation, each unit adds two spaces.
A vector with one entry per gene giving the connectivity of each gene in the weighted network.
Bin Zhang and Steve Horvath (2005) "A General Framework for Weighted Gene Co-Expression Network Analysis", Statistical Applications in Genetics and Molecular Biology: Vol. 4: No. 1, Article 17