full_graph: Generate a similarity network for a list of biclusters

In tdrose/mosbi: Molecular Signature identification using Biclustering

Generate a similarity network for a list of biclusters

Description

The function computes a adjacency matrix for rows and columns of biclusters. The matrix values show, how often two rows or two columns or a row and a column occur together in biclusters. In the resulting adjacency matrix, rows are listed first, followed by columns. They have the same order as the the rows and columns of the input matrix.

Usage

full_graph(
  bics,
  m,
  rr_weight = 1L,
  rc_weight = 1L,
  cc_weight = 1L,
  weighting = 0L
)

Arguments

bics	A list of biclusters.
m	The matrix, that was used to calculated the biclusters.
rr_weight	Weight row-row interactions.
rc_weight	Weight row-col interactions.
cc_weight	Weight col-col interactions.
weighting	Weight interactions by bicluster size. 0 - no weighting, 1 - multiply by bicluster size, 2 - divide by bicluster size.

Details

In case the given biclusters have overall more or less columns than rows, the interactions can be weighted to visualize the result properly.

Value

An adjacency matrix.

Examples

m <- matrix(seq(1:16), nrow=4)
b <- list(bicluster(row=c(1,2,3,4), column=c(1,2,3,4)),
        bicluster(row=c(3,4,5,6), column=c(3,4,5,6)),
        bicluster(row=c(3,4,5,6), column=c(3,4,5,6)))
# full_graph(b, m)

tdrose/mosbi documentation built on May 4, 2022, 3:22 p.m.