runQvalue: Function runQvalue

In jrthompson54/DGE.Tools2: A function library to facilitate Differential Gene Expression Analysis

Description

Takes an list of contrasts (e.g. topTable output or other dataframes that contain a pvalue column). Adds a qvalue and local FDR (lfdr) column to each dataframe.

Usage

runQvalue(contrastList, pvalField = "P.Value", ...)

Arguments

contrastList	An list of dataframes with a pvalue column (all tables must use the same colname for the pval column)
pvalField	Optional. (Default = "P.Value") Not needed if you're working with topTable output. Use to define the colname of the pvalue field in each dataframe.
...	Optional arguments passed to the qvalue function (See ?qvalue)

Details

The qvalue package from John Storey at Princeton takes a list of pvalues and calculates a qvalue and a Local FDR (LFDR). The qvalue is essentially a less conservative FDR estimate compared to the default Benjamini-Hochberg FDR produced by topTable analysis (i.e. will give you more differential genes at the same nominal cutoff). The qvalue function also produces a Local FDR (lfdr) column which answers a slightly different and possibly more relevant question. The BH FDR (adj.P.Val in topTable data.frames) and qvalue tell you what the false discovery rate is for a list of genes at a given threshold. The local FDR attempts to answer the question: what is the probability that this particular gene is a false discovery?

See http://genomine.org/papers/Storey_FDR_2011.pdf for a brief introduction to FDRs and q-values.

Value

The input contrastList now containing qvalue and lfdr columns in each dataframes.

Author(s)

John Thompson, [e-mail]

Examples

MyContrastList <- runQvalue(MyContrastList)
#The magrittr way:
MyContrastList %<>% runQvalue

jrthompson54/DGE.Tools2 documentation built on May 12, 2021, 8:47 p.m.