clusterCt {HTqPCR}R Documentation

Clustering of qPCR Ct values

Description

Hierarchical clustering of samples or genes from high-throughput qPCR experiments, such as the TaqMan Low Density Array platform. Individual clusters can be selected, and the features within them listed in the given order.

Usage

clusterCt(q, main = NULL, type = "genes", dist = "pearson", xlab = "Cluster dendrogram", n.cluster, h.cluster, select.cluster = FALSE, ...)

Arguments

q

object of class qPCRset.

main

character string, plot title.

type

character string, either "genes" (default) or "samples", indicating what is to be clustered.

dist

character string, specifying whether to use "pearson" correlation (default) or "euclidean" distance for the clustering.

xlab

character string, label for the x-axis.

n.cluster

integer, the number of cluster to divide the dendrogram into. See details.

h.cluster

numeric, the height at which to cut the dendrogram into clusters. See details.

select.cluster

logical, whether to select clusters interactively. See details.

...

any other arguments will be passed to the plot function.

Details

This function may be used to cluster the Ct values and present the result as a dendrogram.

The n.cluster and h.cluster parameters are from the rect.hclust function and can be used to divide the dendrogram into subclusters based on either number of clusters or height of branch, drawing boxes around subclusters. The members of each cluster can be returned (see value). If n.cluster is specified h.cluster will be ignored.

If select.cluster is chosen individual subclusters can be selected and marked by a box by clicking on their highest comment branch with the (first) mouse button. Multiple clusters can be selected until any mouse button other than the first is pressed, and the function can be used in conjunction with either n.cluster or h.cluster. The members of each cluster will likewise be returned, in the order they were selected.

Value

A plot is created on the current graphics device. If any subclusters are marked, these will be returned invisibly in a list, with one component for each subcluster. The individual slots in the list contain the names of the genes, and their position in the original input data (row number).

Author(s)

Heidi Dvinge

See Also

hclust, dist, rect.hclust, identify.hclust

Examples

# Load example data
data(qPCRraw)
# Clustering samples
clusterCt(qPCRraw, type="samples")
clusterCt(qPCRraw, type="samples", dist="euclidean")
# Clustering genes
clusterCt(qPCRraw, type="genes", cex=0.5)
clusterCt(qPCRraw, type="genes", h.cluster=1.5, cex=0.5)
cluster.list <- clusterCt(qPCRraw, type="genes", n.cluster=6, cex=0.5)
cluster.list[[1]]

[Package HTqPCR version 1.44.0 Index]