Unified entry point for loading LD data from a metadata TSV file.
Usage
load_LD_matrix(
LD_meta_file_path,
region,
extract_coordinates = NULL,
return_genotype = FALSE,
n_sample = NULL
)Arguments
- LD_meta_file_path
Path to the LD metadata TSV file.
- region
Region of interest: "chr:start-end" string or data.frame with chrom/start/end.
- extract_coordinates
Optional data.frame with columns "chrom" and "pos" for specific coordinates extraction (only for pre-computed LD blocks).
- return_genotype
Controls what LD_matrix contains in the return value. FALSE (default): always return correlation matrix R. TRUE: return genotype matrix X (only valid for PLINK sources). "auto": return X for PLINK sources, R for pre-computed sources.
- n_sample
Optional sample size for computing variance (= 2*p*(1-p)*n/(n-1)). If NULL, ref_panel will not include variance or n_nomiss columns. Only used for PLINK genotype sources.
Value
A list with:
- LD_variants
Character vector of variant IDs (canonical format).
- LD_matrix
LD correlation matrix R (or genotype matrix X when return_genotype is TRUE or "auto" with PLINK source).
- ref_panel
Data.frame with variant metadata (chrom, pos, A2, A1, variant_id, and optionally allele_freq, variance, n_nomiss).
- is_genotype
Logical: TRUE if LD_matrix contains genotype X, FALSE if correlation R.
- block_metadata
Data.frame with region/block info. For pre-computed LD: one row per block. For PLINK: a single row spanning the loaded region.
Details
The metadata TSV must have columns: chrom, start, end, path. Two formats:
Pre-computed LD blocks: many rows per chromosome with block boundaries in start/end and path pointing to .cor.xz files (optionally comma-separated with a .bim path).
PLINK genotype files: one row per chromosome with start=0, end=0, and path pointing to a per-chromosome PLINK prefix (.pgen/.pvar[.zst]/.psam or .bed/.bim/.fam). LD is computed on the fly via
compute_LD().