Prepare Genotype Matrix for LD Computation — prepare_geno • LDxBlocks

Central preparation function called inside Big_LD() and run_Big_LD_all_chr. Depending on method:

"r2": Mean-centres each SNP column. No kinship needed. Returns the centred matrix directly. Fast - O(np).
"rV2": Computes VanRaden GRM via AGHmatrix, tunes it via ASRgenomics, inverts via Cholesky/eigen, and left-multiplies the centred genotype matrix. Returns the whitened matrix and the whitening factor for use in appendSGTs.

Usage

prepare_geno(
  geno,
  method = c("r2", "rV2"),
  kin_method = c("chol", "eigen"),
  verbose = FALSE
)

Arguments

geno: Numeric matrix (individuals x SNPs), 0/1/2.
method: Character. "r2" or "rV2".
kin_method: Character. Whitening decomposition for rV^2: "chol" or "eigen".
verbose: Logical. Print progress.

Value

Named list:

adj_geno: n x p numeric matrix ready for compute_ld().
V_inv_sqrt: n x n whitening matrix, or NULL for r^2.