es_meta: Effect size combination meta analysis.

Description Usage Arguments Details Value Examples

View source: R/meta_utils.R

Description

Performs effect-size meta-analyses across all studies and seperately for each tissue source.

Usage

1
es_meta(diff_exprs, cutoff = 0.3, by_source = FALSE)

Arguments

diff_exprs

Previous result of diff_expr, which can be reloaded using load_diff.

cutoff

Minimum fraction of contrasts that must have measured each gene. Between 0 and 1.

by_source

Should seperate meta-analyses be performed for each tissue source added with add_sources?

Details

Builds on zScores function from GeneMeta by allowing for genes that were not measured in all studies. This implementation also uses moderated unbiased effect sizes calculated by effectsize from metaMA and determines false discovery rates using fdrtool.

Value

A list of named lists, one for each tissue source. Each list contains two named data.frames. The first, filt, has all the columns below for genes present in cutoff or more fraction of contrasts. The second, raw, has only dprime and vardprime columns, but for all genes (NAs for genes not measured by a given contrast).

dprime

Unbiased effect sizes (one column per contrast).

vardprime

Variances of unbiased effect sizes (one column per contrast).

mu

Overall mean effect sizes.

var

Variances of overall mean effect sizes.

z

Overall z score = mu / sqrt(var).

fdr

False discovery rates calculated from column z using fdrtool.

pval

p-values calculated from column z using fdrtool.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
library(lydata)

# location of data
data_dir <- system.file("extdata", package = "lydata")

# gather GSE names
gse_names  <- c("GSE9601", "GSE15069", "GSE50841", "GSE34817", "GSE29689")

# load previous analysis
anals <- load_diff(gse_names, data_dir)

# add tissue sources to perform seperate meta-analyses for each source (optional)
# anals <- add_sources(anals, data_dir)

# perform meta-analysis
es <- es_meta(anals, by_source = TRUE)

crossmeta documentation built on Nov. 8, 2020, 8 p.m.