gosummaries.kmeans: Prepare gosummaries object based on k-means results
In GOsummaries: Word cloud summaries of GO enrichment analysis

Description Usage Arguments Details Value Author(s) Examples

The gosummaries object is created based on the genes in the clusters, it is possible to add corresponding gene expression data as well.

1
2
3

## S3 method for class 'kmeans'
gosummaries(x, exp = NULL, annotation = NULL,
  components = 1:length(x$size), organism = "hsapiens", ...)

`x`	an object of class `kmeans`
`exp`	an expression matrix, with row names corresponding to the names of the genes in clusters (Optional)
`annotation`	a `data.frame` describing the samples, its row names should match with column names of `exp` (Optional)
`components`	numeric vector of clusters to annotate
`organism`	the organism that the gene lists correspond to. The format should be as follows: "hsapiens", "mmusculus", "scerevisiae", etc.
`...`	GO annotation filtering parameters as defined in `gosummaries.default`

The k-means clustering of expression matrix naturally defines a set of gene lists that can be annotated functionally and displayed as a GOsummaries figure. This functon takes in a kmeans object and and converts it to a gosummaries object that can be plotted. If expression matrix is attached then the panel shows the expression values for each gene as boxplots, if not then number of genes is displayed

It is advisable to filter some genes out before doing the clustering since the very large gene lists (more than 2000 genes) might fail the annotation step and are usually not too specific either.

A gosummaries object.

Raivo Kolde <raivo.kolde@eesti.ee>

## Not run: 
data(tissue_example)

# Filter genes and perform k-means
sd = apply(tissue_example$exp, 1, sd)
exp2 = tissue_example$exp[sd > 0.75,]
exp2 = exp2 - apply(exp2, 1, mean)
kmr = kmeans(exp2, centers = 6, iter.max = 100)

# Create gosummaries object  
gs_kmeans = gosummaries(kmr, exp = exp2, annotation = tissue_example$annot)
plot(gs_kmeans, panel_height = 0, components = 1:3, fontsize = 8)
plot(gs_kmeans, classes = "Tissue", components = 1:3, fontsize = 8)

## End(Not run)