image.stamPrediction {stam}R Documentation

Molecular Symptoms Image on StAM Prediction

Description

Shows prediction results for each sample of the class of interest and for all nodes of a StAM model fit as a color coded image.

Usage

image.stamPrediction(x, aclass = NULL, main = NULL, 
                     xlab = NULL, show.graph = TRUE, 
                     max.label.length = 40, 
                     sample.labels = FALSE, full.names = TRUE, 
                     outfile = NULL, ps = FALSE, res = 72, 
                     pointsize = 10, width = 11, 
                     minspec = NULL, minsens = 0.1, 
                     maxsens = 1, invert = FALSE, 
                     col = stam.rgb.colors(r0 = 0), 
                     bin.thresh = NULL, ...)

Arguments

x stamPrediction object to be illustrated
aclass the name of the phenotype class of interest, usually the disease class. If set to NULL the first class in lexicographical order is chosen
main the main title of the plot, generated automatically if left blank
xlab the label of the x-axis
show.graph (default is T) whether or not to show GO relations between the nodes
max.label.length the maximum string length for a GO term
sample.labels whether or not to show sample names on the x-axis
full.names whether or not to show GO terms instead of GO IDs
outfile name of output file if postscript or PNG graphics is to be generated. The extension of the file is chosen automatically according to the ps argument. If no outfile is specified, an interactive plot is attempted
res resolution in points per inch
ps if set to TRUE postscript output is generated
pointsize the standard fontsize
width width of image in inches (the height is computed according to the number of nodes to be shown and the pointsize)
minspec nodes to be shown must be at least this specific. If set to NULL this is chosen such that no more than 50 nodes are shown.
minsens nodes to be shown must be at least this sensitive
maxsens nodes to be shown must be at most this sensitive
invert whether or not to invert background and foreground colors
col the color gradient to code classifier output
bin.thresh threshold for binary color coding, if left at the default NULL, classifier outputs are coded on a continuous scale.
... additional options passed to image

Details

This image illustrates classifier outputs generated during a structured analysis of microarrays (StAM). The centralpart of the image shows the color coded classifier outputs for each sample in the class of interest (columns) and each nodes from StAM's model fit (rows). Nodes can be restricted to those of minimal sensitivity or specificity. Also nodes with particularly high sensitivity can be excluded. The color code for the classifier outputs is shown in a color bar on the right hand side if a continuous is used. If a testset is specified only these samples are used to compute sensitivity and specificity. If the sample names are not displayed on the x-axis, the test samples are marked with capital letters or just vertical bars (bars if there are too many test samples). Sensitivity and specificity is shown to the left of the figure together with the relations between the GO nodes from the Gene Ontology. GO terms are printed on the right of the image and may be used for a clickable map in HTML.

Value

A string to be used on an HTML page to provide a clickable map for the GO terms.

Author(s)

Claudio Lottaz

See Also

stam.predict


[Package stam version 1.0.3 Index]