plotHistDensity {GeneSelectMMD}R Documentation

Plot of histogram and density estimate of the pooled gene expression levels.

Description

Plot of histogram of pooled gene expression levels, composited with density estimate based on the mixture of marginal distributions. The density estimate is based on the assumption that the marginal correlations between subjects are zero.

Usage

plotHistDensity(obj.gsMMD,
                plotFlag="case",
                plotComponent=FALSE,
                myxlab="expression level",
                myylab="density",
                mytitle="Histogram of gene expression levels\nimposed with estimated density (case)",
                x.legend=NULL,
                y.legend=NULL,
                numPoints=500,
                mycol=1:4, 
                mylty=1:4, 
                mylwd=rep(3,4), 
                cex.main=2, 
                cex.lab=1.5, 
                cex.axis=1.5, 
                cex=2,
                bty="n")

Arguments

obj.gsMMD an object returned by gsMMD, gsMMD.default, gsMMD2, or gsMMD2.default
plotFlag logical. Indicate the plot will based on which type of subjects.
plotComponent logical. Indicate if components of the mixture of marginal distribution will be plotted.
myxlab label for x-axis
myylab label for y-axis
mytitle title of the plot
x.legend the x-corrdiates of the legend
y.legend the y-corrdiates of the legend
numPoints logical. Indicate how many genes will be plots.
mycol color for the density estimates (overall and components)
mylty line styles for the density estimates (overall and components)
mylwd line width for the density estimates (overall and components)
cex.main font for main title
cex.lab font for x- and y-axis labels
cex.axis font for x- and y-axis
cex font for texts
bty the type of box to be drawn around the legend. The allowed values are '"o"' and '"n"' (the default).

Details

For a given type of subjects, we pool their expression levels together if the marginal correlations among subjects are zero. We then draw a histogram of the pooled expression levels. Next, we composite density estimates of gene expression levels for the overal distribution and the 3 component distributions.

Value

A list containing coordinates of the density estimates:

x sorted pooled gene expression levels for cases or controls.
x2 a subset of x specified by the sequence: seq(from=1,to=len.x, by=delta), where len.x is the length of the vector x, and delta=floor(len.x/numpoints).
y density estimate corresponding to x2
y1 weighted density estimate for gene cluster 1
y2 weighted density estimate for gene cluster 2
y3 weighted density estimate for gene cluster 3

Note

The density estimate is obtained based on the assumption that the marginal correlation among subjects is zero. If the estimated marginal correlation obtained by gsMMD is far from zero, then do not use this plot function.

Author(s)

Weiliang Qiu stwxq@channing.harvard.edu, Wenqing He whe@stats.uwo.ca, Xiaogang Wang stevenw@mathstat.yorku.ca, Ross Lazarus ross.lazarus@channing.harvard.edu

References

Qiu, W.-L., He, W., Wang, X.-G. and Lazarus, R. (2008). A Marginal Mixture Model for Selecting Differentially Expressed Genes across Two Types of Tissue Samples. The International Journal of Biostatistics. 4(1):Article 20. http://www.bepress.com/ijb/vol4/iss1/20

Examples

  library(ALL)
  data(ALL)
  eSet1 <- ALL[1:100, ALL$BT == "B3" | ALL$BT == "T2"]
  
  mem.str <- as.character(eSet1$BT)
  nSubjects <- length(mem.str)
  memSubjects <- rep(0,nSubjects)
  # B3 coded as 0 (control), T2 coded as 1 (case)
  memSubjects[mem.str == "T2"] <- 1
  
  obj.gsMMD <- gsMMD(eSet1, memSubjects, transformFlag = TRUE, 
    transformMethod = "boxcox", scaleFlag = TRUE, quiet = FALSE)

  plotHistDensity(obj.gsMMD, plotFlag = "case", 
      mytitle = "Histogram of gene expression levels for T2\nimposed with estimated density (case)", 
      plotComponent = TRUE, 
      x.legend = c(0.8, 3), 
      y.legend = c(0.3, 0.4), 
      numPoints = 500)


[Package GeneSelectMMD version 1.0.0 Index]