data_sim: Data Simulation

In songbiostat/wAF: Weighted Adaptive Fisher Method

Description

Data Simulation

Usage

data_sim(
  n,
  K,
  prop,
  strg,
  binary = FALSE,
  case = NULL,
  corr = NULL,
  maf = NULL,
  beta = NULL
)

Arguments

n	Number of subjects.
K	Number of SNVs.
prop	Proportion of true causal/associated SNVs.
strg	Effect size. SNV effects are selected from U(-strg, strg) distribution.
binary	Idicator of whether the trait is binary.
case	Number of cases if binary trait is simulated. Default is n/2.
corr	Correlation matrix of the underlying multivariate normal distribution. Default is a first order autoregressive structure with adjacent correlation 0.9.
maf	Minor allele frequencies for the SNVs.
beta	SNV effects including the constant. If beta is given, K, prop and strg should be missing; otherwise, beta will be overwritten.

Value

A list object.

SNV

A n by K matrix with counts of minor alleles.

trait

A vector with simulated traits for n subjects.

zero_var

Indexes for columns with no variation. These columns should be removed if SNV is further used by perm_score, wAF and wAFd functions.

Examples

# Simulate data using n, K, prop, strg
data1 <- data_sim(100, 10, 0.1, 2)

# Simulate data using given beta
beta0 <- runif(11)
maf <- runif(10, min = 0.001, max = 0.3)
data2 <- data_sim(100, case = 60, maf = maf, beta = beta0)

songbiostat/wAF documentation built on Feb. 26, 2021, 6:24 p.m.