exportRcade-methods: exportRcade and ...

In jmc200/Rcade: R-based analysis of ChIP-seq And Differential Expression - a tool for integrating a count-based ChIP-seq analysis with differential expression summary data

Description

Methods for exporting Rcade objects, either to disk or in R.

Usage

exportRcade(x, directory="RcadeOutput", cutoffMode="top", cutoffArg = 1000, justGeneID=FALSE, removeDuplicates="beforeCutoff")

Arguments

x	An Rcade object.
directory	character - The directory to export output to.
cutoffMode	character - The method to cut off each list (see Details). Must be "all", "top", "B" or "FDR".
cutoffArg	numeric - What cutoff to use (see Details).
justGeneID	logical - if TRUE, export only the geneID column. If FALSE, export all columns.
removeDuplicates	character - Should we remove duplicate GeneIDs and, if so, should we do this before or after applying the cutoff? Must be "beforeCutoff", "afterCutoff" or "none". (If removing duplicates then, for each list, the entry with the highest B value is retained.)

Details

This function exports Rcade output to disk - specifically, it creates the following files:

File:	ChIP:	DE
ChIP.csv	Present (needs log ratio > 0)	Ignored
ChIPonly.csv	Present (needs log ratio > 0)	Absent
DEandChIP.csv	Present (needs log ratio > 0)	Present
DownChIP.csv	Present (needs log ratio > 0)	Present (logFC < 0)
Down.csv	Ignored	Present (logFC < 0)
DownNoChIP.csv	Absent	Present (logFC < 0)
Nothing.csv	Absent	Absent
UpChIP.csv	Present (needs log ratio > 0)	Present (logFC > 0)
Up.csv	Ignored	Present (logFC > 0)
UpNoChIP.csv	Absent	Present (logFC > 0)

Each file contains genes appropriate to its hypothesis, sorted by descending B value (i.e. ranked from most interesting to least interesting). For example, if you wanted the genes that display DE (either up or down) and also have ChIP signal present, you would look at the top rows of DEandChIP.csv. For genes that have a ChIP signal but explicitly show no DE, use ChIPonly.csv.

A cutoff is applied to each list, according to the value of cutoffMode, referring to cutoffArg if necessary:

cutoffMode = "all"

cutoff ignored, all results written to disk.

cutoffMode = "top"

Take the top N genes, where N is specified by cutoffArg.

cutoffMode = "B"

Take all genes with that satisfy B > cutoffArg, where B is the log-odds.

cutoffMode = "FDR"

The expected false positive rate, FPR, and the expected false negative rate, FNR, are calculated using B values.

The cutoff chosen is the one that maximizes the value of FPR + cutoffArg*FNR.

Usage

exportRcade(x, directory="RcadeOutput", cutoffMode="top", cutoff = 100, justGeneID=FALSE, removeDuplicates="beforeCutoff", ...)

Examples

	data(RcadeSTAT1)
	## Not run: exportRcade(RcadeSTAT1)

jmc200/Rcade documentation built on June 17, 2020, 5:59 a.m.