Estimate background paired TCRdist distributions — .estimate_background_tcrdist

For each foreground TCR, estimates the probability of seeing a paired TCRdist score at or below each integer distance from 0 to max_dist under a null model where alpha and beta chains are independently drawn from shuffled backgrounds.

Usage

.estimate_background_tcrdist_distributions(
  organism,
  tcr_df,
  max_dist,
  num_random_samples = 50000L,
  pseudocount = 0.25,
  preserve_vj_pairings = FALSE,
  bg_tcrs = NULL
)

Arguments

organism: Character string. Organism key (e.g. "human").
tcr_df: A data.frame with columns va, ja, cdr3a, cdr3a_nucseq, vb, jb, cdr3b, cdr3b_nucseq.
max_dist: Integer. Maximum paired TCRdist to consider.
num_random_samples: Integer. Number of random background chains to generate per chain type.
pseudocount: Numeric. Pseudocount added before normalizing.
preserve_vj_pairings: Logical. Preserve V-J pairings in resampling.
bg_tcrs: Optional data.frame used for background generation.

Value

Numeric matrix of dimensions nrow(tcr_df) x (max_dist + 1).