Build mr.mash prior covariance matrices — build_mrmash_prior

Shared helper used by both mrmash_wrapper (individual-level) and mrmash_rss_weights (summary statistics). Constructs the S0 list of prior covariance matrices via the canonical mixture (mr.mashr::compute_canonical_covs) and optional data-driven matrices, expanded over a scaling grid via mr.mashr::expand_covs. The prior grid is derived from Bhat and Shat via compute_grid when not supplied.

Usage

build_mrmash_prior_matrices(
  Bhat,
  Shat,
  K = NULL,
  data_driven_prior_matrices = NULL,
  canonical_prior_matrices = TRUE,
  prior_grid = NULL,
  hetgrid = c(0, 0.25, 0.5, 0.75, 1),
  singletons = TRUE
)

Arguments

Bhat: Numeric matrix of effect-size estimates (variants x conditions).
Shat: Numeric matrix of standard errors (variants x conditions).
K: Number of conditions. When NULL, inferred from ncol(Bhat).
data_driven_prior_matrices: Optional list with element U (list of raw covariance matrices) computed e.g. by compute_cov_flash / compute_cov_diag.
canonical_prior_matrices: Logical. When TRUE (default for RSS), include the standard canonical mixture from mr.mashr::compute_canonical_covs(). When FALSE, data_driven_prior_matrices must be supplied.
prior_grid: Optional pre-computed scaling grid (numeric vector). When NULL, derived from Bhat, Shat via compute_grid().
hetgrid: Heterogeneity grid passed to mr.mashr::compute_canonical_covs(). Default c(0, 0.25, 0.5, 0.75, 1), matching the individual-level wrapper.
singletons: Whether to include single-condition prior components. Default TRUE.

Value

A list with components S0 (the expanded list of prior covariance matrices) and prior_grid (the scaling grid that was used).