normalyzerDE: NormalyzerDE differential expression

In NormalyzerDE: Evaluation of normalization methods and calculation of differential expression analysis statistics

Description

Performs differential expression analysis on a normalization matrix. This command executes a pipeline processing the data and generates an annotated normalization matrix and a report containing p-value histograms for each of the performed comparisons.

Usage

normalyzerDE(jobName, comparisons, designPath = NULL, dataPath = NULL,
  experimentObj = NULL, outputDir = ".", logTrans = FALSE,
  type = "limma", sampleCol = "sample", condCol = "group",
  batchCol = NULL, techRepCol = NULL, leastRepCount = 1,
  quiet = FALSE, sigThres = 0.1, sigThresType = "fdr",
  log2FoldThres = 0)

Arguments

jobName	Name of job
comparisons	Character vector containing target contrasts. If comparing condA with condB, then the vector would be c("condA-condB")
designPath	File path to design matrix
dataPath	File path to normalized matrix
experimentObj	SummarizedExperiment object, can be provided as input as alternative to 'designPath' and 'dataPath'
outputDir	Path to output directory
logTrans	Log transform the input (needed if providing non-logged input)
type	Type of statistical comparison, "limma" or "welch"
sampleCol	Design matrix column header for column containing sample IDs
condCol	Design matrix column header for column containing sample conditions
batchCol	Provide an optional column for inclusion of possible batch variance in the model
techRepCol	Design matrix column header for column containing technical replicates
leastRepCount	Minimum required replicate count
quiet	Omit status messages printed during run
sigThres	Significance threshold use for illustrating significant hits in diagnostic plots
sigThresType	Type of significance threshold, "fdr" or "p". "fdr" is strongly recommended (Benjamini-Hochberg corrected p-values)
log2FoldThres	Fold-size cutoff for being considered significant in diagnostic plots

Details

When executed, it performs the following steps:

1: Read the data and the design matrices into dataframes. 2: Generate an instance of the NormalyzerStatistics class representing the data and their statistical comparisons. 3: Optionally reduce technical replicates in both the data matrix and the design matrix 4: Calculate statistical contrats between supplied groups 5: Generate an annotated version of the original dataframe where columns containing statistical key measures have been added 6: Write the table to file 7: Generate a PDF report displaying p-value histograms for each calculated contrast

Value

None

Examples

data_path <- system.file(package="NormalyzerDE", "extdata", "tiny_data.tsv")
design_path <- system.file(package="NormalyzerDE", "extdata", "tiny_design.tsv")
out_dir <- tempdir()
normalyzerDE(
  jobName="my_jobname", 
  comparisons=c("4-5"), 
  designPath=design_path, 
  dataPath=data_path,
  outputDir=out_dir,
  condCol="group")

NormalyzerDE documentation built on Nov. 8, 2020, 8:22 p.m.