mimage: Microarray image

Description Usage Arguments Value Trimming Empty Rows Probe types Examples

View source: R/mimage.r

Description

Visualize microarray probe intensities arranged by their physical location on the array. A false color image is produced for each sample in the microarray object and arranged in a grid.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
mimage(object, type = NULL, select = NULL, colors = NULL,
  legend.label = NULL, nrow = NULL, ncol = NULL, fixed = FALSE,
  empty.rows = "fill", empty.thresh = 0.6, transform = NULL,
  trim = 0.01, fontsize = 12)

## S4 method for signature 'AffyBatch'
mimage(object, type = "pm", select = NULL,
  colors = NULL, legend.label = "Intensity", nrow = NULL, ncol = NULL,
  fixed = FALSE, empty.rows = "fill", empty.thresh = 0.6,
  transform = log2, trim = 0.01, fontsize = 12)

## S4 method for signature 'PLMset'
mimage(object, type = "residuals", select = NULL,
  colors = NULL, legend.label = type, nrow = NULL, ncol = NULL,
  fixed = FALSE, empty.rows = "fill", empty.thresh = 0.6,
  transform = identity, trim = 0.01, fontsize = 12)

## S4 method for signature 'FeatureSet'
mimage(object, type = "pm", select = NULL,
  colors = NULL, legend.label = "Intensity", nrow = NULL, ncol = NULL,
  fixed = FALSE, empty.rows = "fill", empty.thresh = 0.6,
  transform = log2, trim = 0.01, fontsize = 12)

## S4 method for signature 'oligoPLM'
mimage(object, type = "residuals", select = NULL,
  colors = NULL, legend.label = type, nrow = NULL, ncol = NULL,
  fixed = FALSE, empty.rows = "fill", empty.thresh = 0.6,
  transform = identity, trim = 0.01, fontsize = 12)

## S4 method for signature 'array'
mimage(object, type = NULL, select = NULL,
  colors = NULL, legend.label = "Values", nrow = NULL, ncol = NULL,
  fixed = FALSE, empty.rows = "ignore", empty.thresh = 1,
  transform = identity, trim = 0, fontsize = 12)

Arguments

object

a valid Bioconductor microarray data structure

type

for microarray objects type refers to probe type; for objects containing probe-level models (e.g., PLMsets) type refers to the value type (i.e, "residuals" or "weights"). See probe type section for more information.

select

a numeric, character or logical vector indicating samples to include

colors

a vector of colors used to represent probe values

legend.label

Legend label

nrow

optional, number of rows in grid layout

ncol

optional, number of columns in grid layout

fixed

Force images to assume a fixed aspect ratio corresponding to their physical dimensions

empty.rows

Should empty rows be filled with values from neighboring rows (the default, "fill"), should they be dropped ("drop") entirely, or should they be left alone ("ignore")

empty.thresh

what proportion of features must be missing from a row to consider that row empty

transform

a function to be applied to the values prior to visualizatio

trim

a percentile (default = 0.02) or range or 2 values see trimming section for details

fontsize

font size for labels and legend

Value

invisibly a gtable matrix of grobs

Trimming

By default, a 98% winsorization is performed prior to visualization, pulling in values outside of the 1st and 99th percentiles to their respective endpoints. This can be modified using the trim argument to provide either a new percentile, or a range of 2 values defining the min/max of the trimmed endpoints. Set trim = 0 to avoid trimming entirely.

Empty Rows

As a result of platform design, the presence unindexed probes or probe selection by the user (e.g., including only "pm" probes), the matrix-representation of a microarray can contain numerous rows comprised entirely (or mostly) of missing values, which may produce undesirable rasterization artifacts in the microarray image. To avoid this, empty rows are filled with values from a neighboring row. The threshold for what constitutes an empty row can be tweaked with the empty.thresh argument.

Probe types

For microarray data structures the type argument determines the type of probe that should be included. The following table provides a list of valid values for each supported microarray class:

AffyBatch "all" "pm" "mm" -
ExpressionFeatureSet "all" "pm" "mm" -
GeneFeatureSet "all" "pm" - "bg"
ExonFeatureSet "all" "pm" "mm" "bg"
SnpFeatureSet "all" "pm" "mm" -

Examples

1
2
3
4
5
6
7
8
# standard array visualization
mimage(iris3)

# microarray visualization
if (require(affydata, quietly = TRUE)) {
  data("Dilution", package = "affydata")
  mimage(Dilution, select = c("20A", "10A"))
}

Bioconductor-mirror/mimager documentation built on June 1, 2017, 12:44 p.m.