pcaExplorer (version 1.0.2)

pca2go: Functional interpretation of the principal components

Description

Extracts the genes with the highest loadings for each principal component, and performs functional enrichment analysis on them using routines and algorithms from the topGO package

Usage

pca2go(se, pca_ngenes = 10000, annotation = NULL,
  inputType = "geneSymbol", organism = "Mm", ensToGeneSymbol = FALSE,
  loadings_ngenes = 500, background_genes = NULL, scale = FALSE, ...)

Arguments

se

A DESeqTransform object, with data in assay(se), produced for example by either rlog or varianceStabilizingTransformation

pca_ngenes

Number of genes to use for the PCA

annotation

A data.frame object, with row.names as gene identifiers (e.g. ENSEMBL ids) and a column, gene_name, containing e.g. HGNC-based gene symbols

inputType

Input format type of the gene identifiers. Will be used by the routines of topGO

organism

Character abbreviation for the species, using org.XX.eg.db for annotation

ensToGeneSymbol

Logical, whether to expect ENSEMBL gene identifiers, to convert to gene symbols with the annotation provided

loadings_ngenes

Number of genes to extract the loadings (in each direction)

background_genes

Which genes to consider as background.

scale

Logical, defaults to FALSE, scale values for the PCA

...

Further parameters to be passed to the topGO routine

Value

• A nested list object containing for each principal component the terms enriched in each direction. This object is to be thought in combination with the displaying feature of the main pcaExplorer function

Examples

library(airway)
library(DESeq2)
data(airway)
airway
dds_airway <- DESeqDataSet(airway, design= ~ cell + dex)
rld_airway <- rlogTransformation(dds_airway)

# constructing the annotation object
anno_df <- data.frame(gene_id = rownames(dds_airway),
                      stringsAsFactors=FALSE)
library("AnnotationDbi")
library("org.Hs.eg.db")
anno_df$gene_name <- mapIds(org.Hs.eg.db,
                            keys=anno_df$gene_id,
                            column="SYMBOL",
                            keytype="ENSEMBL",
                            multiVals="first")
rownames(anno_df) <- anno_df$gene_id
bg_ids <- rownames(dds_airway)[rowSums(counts(dds_airway)) > 0]
library(topGO)
pca2go_airway <- pca2go(rld_airway,
                        annotation = anno_df,
                        organism = "Hs",
                        ensToGeneSymbol = TRUE,
                        background_genes = bg_ids)