chroGPS (version 1.16.0)

profileClusters: Compute enrichment/depletion ratio for the observed epigenetic profiles in epigenetic clusters.

Description

The function computes the ratio between the proportion of epigenetic mark presence in the clusters given as input and that observed for all elements. Results are returned as a numerical matrix, easily visualized in the shape of a classical heatmap.

Usage

profileClusters(x, uniqueCount = TRUE, weights, clus, i, minpoints, merged = FALSE, log2 = TRUE, plt = FALSE)

Arguments

x

Genes * Factors matrix or data frame used for generating epigene clusters, indicating 1 for binding of factor j in gene i, 0 otherwise.

uniqueCount

Logical value to indicate if clusters come from epigenes (identical rows in x are merged into a single one) or genes (every row in x is mantained). See help for uniqueCount for details.

weights

Named vector analog to that on distGPS. Names are used as unique column names (i.e. epigenetic factors) so that enrichment profiles for replicates of the same epigenetic factor can be merged into a single element by computing its average enrichment (arithmetic mean).

clus

clusGPS object with epigenetic clusters generated from pairwise distances from x, as generated by the clusGPS function. See help for distGPS and clusGPS for details.

i

Clustering entry from which cluster profiling is to be computed.

minpoints

(Optional). Ignore clusters with fewer than minpoints, deprecated.

merged

(Optional). If clusters provided have been previously merged or not, deprecated.

log2

Logical to indicate if enrichment/depletion proportions are returned in log2 scale. Defaults to TRUE.

plt

Deprecated.

Value

A numerical matrix with the enrichment/depletion profile of the epigenetic marks for each cluster provided in the clusGPS object. Easy to visualize for instance with a heatmap plot.

See Also

distGPS for computing pairwise distances between epigenetic elements. clusGPS for computing epigenetic clusters.

Examples

# Not run
# data(s2)
# # Computing distances
# d <- distGPS(s2.tab,metric='tanimoto',uniqueRows=TRUE)
# # Creating MDS object
# mds1 <- mds(d,type='isoMDS')
# mds1
# plot(mds1)
# Precomputing clustering
# h <- hclust(as.dist(d@d),method='average')
# # Calculating densities (contours and probabilities), takes a while
# clus <- clusGPS(d,mds1,preMerge=TRUE,k=max(cutree(h,h=0.5)))
# Computing cluster profiles
# p1 <- profileClusters(s2.tab, uniqueCount = TRUE, clus, i=125, minpoints=30, log2 = TRUE, plt = FALSE)
# Requires gplots
# heatmap.2(p1,col=redblue(100))