subsampleDb {tigger}

Subsample repertoire

Description

subsampleDb will sample the same number of sequences for each gene, family or allele (specified with mode) in data. Samples or subjects can be subsampled indepently by setting group.

Usage

subsampleDb(
  data,
  gene = "v_call",
  mode = c("gene", "allele", "family"),
  min_n = 1,
  max_n = NULL,
  group = NULL
)

Arguments

data

data.frame containing repertoire data.

gene

name of the column in data with allele calls. Default is v_call.

mode

one of c("gene", "family", "allele") defining the degree of specificity regarding allele calls when subsetting sequences. Determines how data will be split into subsets from which the same number of sequences will be subsampled. See also group.

min_n

minimum number of observations to sample from each groupe. A group with less observations than the minimum is excluded.

max_n

maximum number of observations to sample for all mode groups. If NULL, it will be set automatically to the size of the smallest group. If max_n is larger than the availabe number of sequences for any mode group, if will be automatically adjusted and the efective max_n used will be the size of the smallest mode group.

group

columns containing additional grouping variables, e.g. sample_id. These groups will be subsampled independently. If max_n is NULL, a max_n will be automatically set for each group.

Details

data will be split into gene, allele or family subsets (mode) from which the same number of sequences will be subsampled. If mode=gene, for each gene in the field gene from data, a maximum of max_n sequences will be subsampled. Input sequences that have multiple gene calls (ties), can be subsampled from any of their calls, but these duplicated samplings will be removed, and the final subsampled data will contain unique rows.

Value

Subsampled version of the input data.

See Also

selectNovel

Examples

subsampleDb(AIRRDb)