Flexible partitioning of heritability — partition

An R implementation of the LD score regression method to estimate SNP heritability, focusing on partioning heritability across multiple annotations.

Usage

partition_h2(
  sumstat,
  ldscore_dirs,
  subset_annots = NULL,
  overlapping_annotations = FALSE,
  weights = NULL,
  n_blocks = 200
)

sumstat: A dplyr::tibble() with columns SNP, Z and N
ldscore_dirs: a list of directories containing ldscore files
subset_annots: a character vector of annotations to subset the ldscore files. Use get_annot_names()
overlapping_annotations: are the annotations overlapping? In such a case, the estimate of the enrichment estimate needs to be adjusted.
weights: Optional, a data.frame or tbl with columns SNP, L2
n_blocks: Number of blocks to use for the jackknife estimator

a tibble with results

if (FALSE) { # \dontrun{
partition_h2(sumstat, ldscore_dirs, subset_annots = c("L2", "L3"))
} # }