validatedSignatures: Validation result in data frame

In shbrief/GenomicSuperSignature: Interpretation of RNA-seq experiments through robust, efficient comparison to public databases

Validation result in data frame

Description

Validation result in data frame

Usage

validatedSignatures(
  val_all,
  RAVmodel,
  num.out = 5,
  scoreCutoff = NULL,
  swCutoff = NULL,
  clsizeCutoff = NULL,
  indexOnly = FALSE,
  whichPC = NULL,
  filterMessage = TRUE
)

Arguments

val_all	An output matrix from validate function. If this input is from multiple datasets, only scoreCutoff argument will be considered and other inputs will be ignored.
RAVmodel	PCAGenomicSignatures-class object. RAVmodel used to prepare val_all input.
num.out	A number of highly validated RAVs to output. Default is 5. If any of the cutoff parameters are provided, num.out or the number of filtered RAVs, whichever smaller, will be chosen.
scoreCutoff	A numeric value for the minimum correlation. For multi-studies case, the default is 0.7.
swCutoff	A numeric value for the minimum average silhouette width.
clsizeCutoff	An integer value for the minimum cluster size.
indexOnly	A logical. Under the default (= FALSE), the detailed information on validated RAVs, such as score, average silhouette width, cluster size, is printed. If it is set TRUE, only the RAV number will be printed.
whichPC	An integer value between 1 and 8. PC number of your data to check the validated signatures with. Under the default (NULL), it outputs top scored signatures with any PC of your data.
filterMessage	A logical. Under the default TRUE, any output RAV belong to the filtering list will give a message. Silence this message with filterMessage=FALSE. You can check the filter list using data("filterList").

Value

A subset of the input matrix, which meets the given condition.

Examples

data(miniRAVmodel)
library(bcellViper)
data(bcellViper)
val_all <- validate(dset, miniRAVmodel)
validatedSignatures(val_all, miniRAVmodel, num.out = 3, scoreCutoff = 0)

shbrief/GenomicSuperSignature documentation built on May 3, 2023, 10:07 p.m.