Skip to contents

Join two TCR datasets by TCRdist threshold

Source: R/tcr_join.R
tcrdist_join.Rd

Performs a fuzzy merge of two TCR data.frames: for each row in left_df, finds rows in right_df within radius TCRdist units. Similar to a SQL JOIN but using distance instead of exact matching.

Usage

tcrdist_join(
  left_df,
  right_df,
  organism,
  radius,
  max_n = 5L,
  type = c("inner", "left"),
  suffix = c("_x", "_y"),
  rect_dist_matrix = NULL
)

Arguments

left_df

Data.frame with TCR columns (va, vb, cdr3a, cdr3b) plus any extra columns.

right_df

Data.frame with TCR columns (va, vb, cdr3a, cdr3b) plus any extra columns.

organism

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

radius

Numeric. Maximum TCRdist for a match.

max_n

Integer. Maximum number of matches per left row. Default 5L. Closest matches are kept.

type

Character string. Join type: "inner" (default) keeps only matched pairs, "left" keeps all left rows (NAs for unmatched).

suffix

Character vector of length 2. Suffixes for disambiguating column names. Default c("_x", "_y").

rect_dist_matrix

Optional precomputed rectangular distance matrix (rows = left, cols = right). If provided, organism is not used for distance computation.

Value

A data.frame with columns from both sides (suffixed if overlapping) plus a tcrdist column.

See also

tcrdist_rect, tcrdist_radius_neighbors

Examples

if (FALSE) { # \dontrun{
matches <- tcrdist_join(query_tcrs, reference_tcrs, "human", radius = 50)
} # }