diffAnalyses: Perform statistical analyses

In psichomics: Graphical Interface for Alternative Splicing Quantification, Analysis and Visualisation

Description

Perform statistical analyses

Usage

diffAnalyses(
  data,
  groups = NULL,
  analyses = c("wilcoxRankSum", "ttest", "kruskal", "levene", "fligner"),
  pvalueAdjust = "BH",
  geneExpr = NULL,
  inputID = "sparklineInput"
)

Arguments

data	Data frame or matrix: gene expression or alternative splicing quantification
groups	Named list of characters (containing elements belonging to each group) or character vector (containing the group of each individual sample); if NULL, sample types are used instead when available, e.g. normal, tumour and metastasis
analyses	Character: statistical tests to perform (see Details)
pvalueAdjust	Character: method used to adjust p-values (see Details)
geneExpr	Character: name of the gene expression dataset (only required for density sparklines available in the interactive mode)
inputID	Character: identifier of input to get attributes of clicked event (Shiny only)

Details

The following statistical analyses may be performed simultaneously via the analysis argument:

• ttest - Unpaired t-test (2 groups)

• wilcoxRankSum - Wilcoxon Rank Sum test (2 groups)

• kruskal - Kruskal test (2 or more groups)

• levene - Levene's test (2 or more groups)

• fligner - Fligner-Killeen test (2 or more groups)

• density - Sample distribution per group (only usable through the visual interface)

The following p-value adjustment methods are supported via the pvalueAdjust argument:

• none: do not adjust p-values

• BH: Benjamini-Hochberg's method (false discovery rate)

• BY: Benjamini-Yekutieli's method (false discovery rate)

• bonferroni: Bonferroni correction (family-wise error rate)

• holm: Holm's method (family-wise error rate)

• hochberg: Hochberg's method (family-wise error rate)

• hommel: Hommel's method (family-wise error rate)

Value

Table of statistical analyses

See Also

Other functions to perform and plot differential analyses: plotDistribution()

Examples

# Calculate PSI for skipped exon (SE) and mutually exclusive (MXE) events
eventType <- c("SE", "MXE")
annot <- readFile("ex_splicing_annotation.RDS")
junctionQuant <- readFile("ex_junctionQuant.RDS")

psi <- quantifySplicing(annot, junctionQuant, eventType=c("SE", "MXE"))
group <- c(rep("Normal", 3), rep("Tumour", 3))
diffAnalyses(psi, group)

psichomics documentation built on Nov. 8, 2020, 5:44 p.m.