Create correlation matrix — createCorrelationMatrix • decorate

Create correlation matrix based on correlation between pairs of peaks

Usage

createCorrelationMatrix(
  query,
  regionQuant,
  adjacentCount = 500,
  windowSize = 1e+06,
  method = "adjacent",
  method.corr = c("pearson", "spearman"),
  quiet = FALSE,
  setNANtoZero = FALSE
)

Arguments

query: GRanges object of intervals to query
regionQuant: normalized quantifications of regions in query. Rows are features, like in limma
adjacentCount: number of adjacent entries to compute correlation against
windowSize: width of window in bp around each interval beyond which weight is zero
method: 'adjacent': compute corr on fixed count sliding window define by adjacentCount. "distance": compute corr for entries within windowSize bp
method.corr: specify which correlation method: "pearson" or "spearman"
quiet: suppress messages
setNANtoZero: replace NAN correlation values with a zero

Value

for peak i and j with distance d_i,j, M[i,j] = cor( vobj$E[i,], vobj$E[j,] )

return sparse symmatric matrix

Examples


data('decorateData')

C = createCorrelationMatrix(simLocation, simData)
#> adjacent: 500entries
#> Covariance matrix...
#>  sparsity:49.888 %
#>  memory usage:1.2 Mb