Identify quasi-public meta-clonotypes

Finds TCR sequences that are shared across multiple subjects (individuals) based on TCRdist neighborhoods. A meta-clonotype is defined by a center TCR and a radius: all TCRs within the radius are considered part of the meta-clonotype.

Usage

find_meta_clonotypes(
  tcr_df,
  organism,
  radius = NULL,
  background_df = NULL,
  ctrl_bkgd = 1e-05,
  max_radius = 50L,
  min_nsubject = 2L,
  subject_col = "subject",
  dist_matrix = NULL
)

Arguments

tcr_df

Data.frame with TCR columns (va, vb, cdr3a, cdr3b) and a subject column.

organism

Character string ("human" or "mouse").

radius

Numeric or integer vector. TCRdist radius for neighborhood definition. If a single value, used for all TCRs. If NULL, per-TCR radii are computed from background ECDF.

background_df

Data.frame. Background TCRs for radius computation. Required if radius is NULL.

ctrl_bkgd

Numeric. Background proportion threshold for automatic radius. Default 1e-5.

max_radius

Numeric. Maximum radius to consider. Default 50L.

min_nsubject

Integer. Minimum number of distinct subjects in a neighborhood. Default 2L.

subject_col

Character string. Column name for subject IDs. Default "subject".

dist_matrix

Optional precomputed distance matrix. If provided, tcr_df and organism are not used for distance computation.

Value

A data.frame with one row per meta-clonotype center:

center_index

Row index in tcr_df.

va, cdr3a, vb, cdr3b

Center TCR sequence.

radius

TCRdist radius used.

K_neighbors

Total neighbors within radius.

nsubject

Number of distinct subjects.

neighbor_indices

Comma-separated neighbor row indices.

See also

summarize_meta_clonotype, find_clumping

Examples

if (FALSE) { # \dontrun{
meta <- find_meta_clonotypes(tcr_df, "human", radius = 30,
                              subject_col = "subject")
} # }