Convert a redeemR variant & depth matrix into long format for downstream analysis

This function filters a variant count matrix (mat_var) to include only cells in the provided whitelist and variants observed in at least two cells, applies the same filtering to the corresponding depth matrix (mat_depth), and then melts both into a single long-format data.table. The output contains one row per cell–variant pair with its count (a) and depth (d).

Usage

convert_redeem_matrix_long(
  mat_var,
  mat_depth,
  cell_whitelist = NULL,
  sample = NULL
)

Arguments

mat_var

A numeric matrix of variant UMI counts (cells × variants).

mat_depth

A numeric matrix of total depth (same dimensions and row/column names as mat_var).

cell_whitelist

A character vector of cell barcodes to retain (rows of mat_var/mat_depth).

sample

Character; an identifier for this dataset (printed to the console).

Value

A data.table with columns:

• cell: cell barcode

• variant: variant identifier

• a: UMI count (filtered)

• d: corresponding depth

Details

Included in redeemR-2.0 2025-07-01

1. Retains only rows in mat_var whose rownames appear in cell_whitelist.

2. Drops variants observed in fewer than two cells.

3. Ensures both matrices share the same dimensions before melting.

4. Returns a combined long table for joint count/depth analysis.