meta_analyze: Meta-analyze trcQTL and ascQTL results

In liangyy/mixqtl: Combine total and allele-specific counts for cis-QTL analysis

Description

For each variant/gene pair, meta-analyze summary statistics from trcQTL and ascQTL. If sample size is greater than or equal to n_cutoff, approximate test statistic as z-score, otherwise compute p-value from student-t distribution with d.o.f. = 1. If both trcQTL and ascQTL have no fewer than n_cutoff samples, perform inverse variance-based meta-analysis.

Usage

meta_analyze(trc, asc, n_cutoff = 15)

Arguments

trc	list of effect size estimates and standard deviation (e.g. output of matrix_ls_trc)
asc	list of effect size estimates and standard deviation (e.g. output of matrix_ls_asc)
n_cutoff	threshold to approximate test statistic as z-score.

Value

a list of summary statistics and p-values obtained from trc, asc, and meta-analysis of the two the object within list is named by the method (trc, asc, and meta) and each of which includes pval (p-value), stat (test statistic), stat_type (z-value or t-value), bhat (effect size estimate), se (standard deviation of effect size), method (only for 'meta' object showing if the result is from 'trc', 'asc', or 'meta')

Examples

meta_analyze(
  trc = list(
    beta_hat = matrix(rnorm(100), ncol = 20),
    beta_se = abs(matrix(rnorm(100), ncol = 20)),
    sample_size = 100
  ),
  asc = list(
    beta_hat = matrix(rnorm(100), ncol = 20),
    beta_se = abs(matrix(rnorm(100), ncol = 20)),
    sample_size = 100
  )
)

liangyy/mixqtl documentation built on Sept. 17, 2020, 11:36 a.m.