Meta-analysis Z statistic

Share:

Description

The function calculates a meta analysis Z statistic based on an input data frame of Z statistics.

Usage

1
metaZfunction(datZ, columnweights = NULL)

Arguments

datZ

Matrix or data frame of Z statistics (assuming standard normal distribution under the null hypothesis). Rows correspond to genes, columns to independent data sets.

columnweights

optional vector of non-negative numbers for weighing the columns of datZ.

Details

For example, if datZ has 3 columns whose columns are labelled Z1,Z2,Z3 then ZMeta= (Z1+Z2+Z3)/sqrt(3). Under the null hypothesis (where all Z statistics follow a standard normal distribution and the Z statistics are independent), ZMeta also follows a standard normal distribution. To calculate a 2 sided p-value, one an use the following code pvalue=2*pnorm(-abs(ZMeta) )

Value

Vector of meta analysis Z statistic. Under the null hypothesis this should follow a standard normal distribution.

Author(s)

Steve Horvath

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.