Skip to contents

Annotate consensus segments across cluster with CNA states

Source: R/cnaCalling.R
annotateSegments.Rd

This function annotates consensus segments across clusters using segment-level data from each cluster. It assigns states (i.e., gain, loss, cnloh, neutral), computes proportions of cells per cluster, and defines whether alterations are clonal.

Usage

annotateSegments(
  x,
  consensus_segs,
  ncells,
  ploidy = 2,
  minoverlap = 1e+06,
  clonal.thresh = 0.9
)

Arguments

x

A data frame containing cluster segments information with the following required columns:

  • chrom: Chromosome name (factor or character).

  • start: Start position of the segment (numeric).

  • end: End position of the segment (numeric).

  • cluster: Cluster identifier (numeric or integer).

  • tcn.em: Total copy number (EM algorithm) (numeric).

  • lcn.em: Lower copy number (EM algorithm) (numeric).

consensus_segs

A data.frame of unique consensus segments.

ncells

A named vector specifying the number of cells per cluster (numeric vector). The names must match the cluster identifiers in the cluster column of x.

ploidy

An integer specifying the expected ploidy (default is 2). Used to determine whether a segment is classified as gain, loss, or neutral.

minoverlap

A non-negative integer specifying the minimum overlap between a cluster specific segment and a consensus segment (see IRanges::findOverlaps()) (integer). Default: 1e6.

clonal.thresh

Minimum cell proportion to label a segment as clonal. Default: 1e6.

Value

A data.frame with the consensus segments annotated per cluster, including CNA state, clonal status, and proportion of cells per cluster and across clusters:

  • ncells: Number of cells in cluster.

  • prop.cluster: Proportion of cells in cluster relatively to the total number of cells.

  • gnl: GNL (gain ; neutral ; loss) status as an integer value (1 ; 0 ; -1).

  • loh: Loss of heterozygosity status (logical).

  • state: State of segment (gain ; loss ; neu ; cnloh).

  • cna: Segment CNA status (logical).

  • cna_state: State of CNA segment (gain ; loss ; cnloh).

  • prop.tot: Proportion of cells for the state across clusters.

  • state_clonal: Clonal state of segment (gain ; loss ; neu ; cnloh).

  • cna_clonal: Segment clonal CNA status (logical).

  • cna_clonal_state: Clonal state of CNA segment (gain ; loss ; cnloh).

See also

getSegConsensus(), cnaCalling().

Examples

#' # Example data frame
segs <- data.frame(
    chrom = c("chr1", "chr1", "chr2", "chr2"),
    start = c(1.2e6, 1.1e6, 3.1e6, 3.2e6),
    end = c(2.5e6, 2.6e6, 5.5e6, 5.7e6),
    cluster = c("1", "2", "1", "2"),
    cf.em = c(1, 1, 1, 1),
    tcn.em = c(1, 2, 3, 3),
    lcn.em = c(0, 1, 1, 1)
)

consensus_segs <- getSegConsensus(
    x = segs,
    ncells = c("1" = 50, "2" = 30)
)

# Annotate consensus segments
table <- annotateSegments(
    x = segs,
    consensus_segs = consensus_segs,
    ncells = c("1" = 50, "2" = 30)
)

print(table)
#>   chrom   start     end id cluster cf.em tcn.em lcn.em ncells prop.cluster gnl
#> 1  chr1 1200000 2500000  1       1     1      1      0     50        0.625  -1
#> 2  chr2 3100000 5500000  2       1     1      3      1     50        0.625   1
#> 3  chr1 1200000 2500000  1       2     1      2      1     30        0.375   0
#> 4  chr2 3100000 5500000  2       2     1      3      1     30        0.375   1
#>     loh state   cna cna_state prop.tot state_clonal cna_clonal cna_clonal_state
#> 1  TRUE  loss  TRUE      loss    0.625         <NA>      FALSE             <NA>
#> 2 FALSE  gain  TRUE      gain    1.000         gain       TRUE             gain
#> 3 FALSE   neu FALSE      <NA>    0.375         <NA>      FALSE             <NA>
#> 4 FALSE  gain  TRUE      gain    1.000         gain       TRUE             gain