Get Copy Number Sequence Similarity or Distance Matrix

Get Copy Number Sequence Similarity or Distance Matrix

get_score_matrix(
  x,
  sub_mat = NULL,
  simple_version = FALSE,
  block_size = NULL,
  dislike = FALSE,
  cores = 1L,
  verbose = FALSE
)

Arguments

x	a coding copy number sequence (valid letters are A to X).
sub_mat	default is NULL, use longest common substring method. It can be a substitution matrix, each element indicates a score to plus. See build_sub_matrix().
simple_version	if TRUE, just use segmental copy number value.
block_size	a block size to aggregrate, this is designed for big data, it means results from adjacent sequences will be aggregrate by means to reduce the size of result matrix.
dislike	if TRUE, returns a dissimilarity matrix instead of a similarity matrix.
cores	computer cores, default is 1, note it is super fast already, set more cores typically do not speed up the computation.
verbose	if TRUE, print extra message, note it will slower the computation.

Value

a score matrix.

Examples

load(system.file("extdata", "toy_segTab.RData",
  package = "CNVMotif", mustWork = TRUE
))
x <- transform_seqs(segTabs)
x
seqs <- extract_seqs(x$dt)
seqs
seqs2 <- extract_seqs(x$dt, flexible_approach = TRUE)
seqs2

mat <- get_score_matrix(seqs$keep, x$mat, verbose = TRUE)
mat

mat2 <- get_score_matrix(seqs$keep, x$mat, dislike = TRUE)
identical(mat2, 120L - mat)

mat_b <- get_score_matrix(seqs$keep, x$mat, block_size = 2L)
## block1 represents the first 2 sequences
## block2 represents the 3rd, 4th sequences
## ...
mat_b

mat_c <- get_score_matrix(seqs$keep)
mat_c
mat_d <- get_score_matrix(seqs$keep, dislike = TRUE)
mat_d
# \donttest{
if (requireNamespace("doParallel")) {
  mock_seqs <- sapply(1:10000, function(x) {
    paste(sample(LETTERS[1:24], 5, replace = TRUE), collapse = "")
  })

  system.time(
    y1 <- get_score_matrix(mock_seqs, x$mat, cores = 1)
  )

  system.time(
    y2 <- get_score_matrix(mock_seqs, x$mat, cores = 2)
  )

  all.equal(y1, y2)
}
# }