Repertoire overlap metrics

Computes overlap metrics between two named count vectors representing clonotype abundances from two samples. Available metrics:

Usage

tcr_repertoire_overlap(
  counts_a,
  counts_b,
  metrics = c("jaccard", "morisita_horn", "overlap_coef")
)

Arguments

counts_a, counts_b

Named numeric vectors of clonotype counts. Names are clonotype identifiers.

metrics

Character vector. Which metrics to compute. Default computes all three.

Value

A named list with the requested metrics, each a scalar in [0, 1].

Details

Jaccard index

|A intersect B| / |A union B| — proportion of shared clonotype species (presence/absence).

Morisita-Horn

2 * sum(p_a * p_b) / (sum(p_a^2) + sum(p_b^2)) — abundance-weighted overlap.

Overlap coefficient

|A intersect B| / min(|A|, |B|) — overlap normalized by the smaller set.

See also

tcr_diversity, tcr_clonality

Examples

a <- c(clone1 = 10, clone2 = 5, clone3 = 1)
b <- c(clone1 = 8, clone3 = 3, clone4 = 2)
tcr_repertoire_overlap(a, b)
#> $jaccard
#> [1] 0.5
#> 
#> $morisita_horn
#> [1] 0.8420231
#> 
#> $overlap_coef
#> [1] 0.6666667
#>