Fine-mapping using an inverse-moment prior.

Usage

finimom(
  beta,
  se,
  eaf,
  R,
  n = NULL,
  cs = TRUE,
  cs_level = 0.95,
  pip = FALSE,
  maxsize = 10,
  tau = NULL,
  r = 1,
  niter = 12500,
  burnin = 2500,
  seed = 456,
  excl.burnin = TRUE,
  a0 = 1,
  b0 = NULL,
  u = NULL,
  inds0 = NULL,
  standardize = TRUE,
  verbose = TRUE,
  insampleLD = NULL,
  clump = TRUE,
  clump_r2 = 0.99^2,
  check_ld = FALSE,
  ala = NULL,
  purity = NULL
)

Arguments

beta: Effect size estimates.
se: Standard errors.
eaf: Effect allele frequencies.
R: LD-matrix.
n: GWAS sample size. If provided, then tau is calculated based on n.
cs: Are credible sets returned?
cs_level: Credible set level.
pip: Are posterior inclusion probabilities returned?
maxsize: Maximum model size.
tau: Prior parameter tau.
r: Prior parameter r.
niter: Number of iterations.
burnin: Number of burn-in iterations.
seed: Random seed.
excl.burnin: Should burn-in be excluded?
a0: Hyperparameter a for the model size prior.
b0: Hyperparameter b for the model size prior (but preferably use parameter u).
u: Hyperparameter for model size prior. Defaults to 2 for in-sample LD matrix and 2.25 for out-of-sample LD matrix.
inds0: Indices for the starting model.
standardize: Should the effect sizes be standardised? Defaults to TRUE.
verbose: Verbose output.
insampleLD: Is in-sample LD used?
clump: Should clumping be done for extremely highly-correlated variants?
clump_r2: Clumping threshold for extremely highly-correlated variants.
check_ld: Should LD discrepancy check be performed?
ala: Whether Approximate Laplace should be used?
purity: Credible set purity, defined as the minimum absolute correlation between the variants in a credible set.

Value

List.