GMMAT (version 1.4.2)

glmm.score.meta: Performing meta-analysis for GLMM based score test results

Description

Use output files from GLMM based score tests to perform meta-analysis.

Usage

glmm.score.meta(files, outfile, SNP = rep("SNP", length(files)), 
	A1 = rep("A1", length(files)), A2 = rep("A2", length(files)))

Value

a data frame containing the following:

SNP

SNP name.

A1

allele 1.

A2

allele 2.

N

total sample size.

AF

effect allele frequency (user-defined: can be either allele 1 or allele 2).

SCORE

the summary score of the effect allele.

VAR

the variance of the summary score.

PVAL

meta-analysis p-value.

Arguments

files

a vector of input file names. The input files should be the output files of glmm.score( ), or customized tab or space delimited files that include at least 8 columns: SNP, effect allele, noneffect allele, N, AF, SCORE, VAR and PVAL. The column names of SNP, effect allele and noneffect allele can be customized and provided in SNP, A1 and A2.

outfile

the output file name.

SNP

a character vector of SNP column names in each input file. The length and order must match the length and order of files (default = rep("SNP", length(files))).

A1

a character vector of allele 1 column names in each input file. The length and order must match the length and order of files (default = rep("A1", length(files))). Note that glmm.score.meta( ) does not define A1 as the effect allele or noneffect allele: it is the user's choice. However, the choice should be consistent across different studies, if A1 column is the effect allele in one study but the noneffect allele in another, meta-analysis results will be incorrect.

A2

a character vector of allele 2 column names in each input file. The length and order must match the length and order of files (default = rep("A2", length(files))). Note that glmm.score.meta( ) does not define A2 as the effect allele or noneffect allele: it is the user's choice. However, the choice should be consistent across different studies, if A2 column is the effect allele in one study but the noneffect allele in another, meta-analysis results will be incorrect.

Author

Han Chen

See Also

glmm.score

Examples

# \donttest{
infile1 <- system.file("extdata", "meta1.txt", package = "GMMAT")
infile2 <- system.file("extdata", "meta2.txt", package = "GMMAT")
infile3 <- system.file("extdata", "meta3.txt", package = "GMMAT")
outfile <- tempfile()
glmm.score.meta(files = c(infile1, infile2, infile3), outfile = outfile,
	SNP = rep("SNP", 3), A1 = rep("A1", 3), A2 = rep("A2", 3))
unlink(outfile)
# }