plotCat: Plotting correspondence at the top curves

In marchion/matchBox: Utilities to compute, compare, and plot the agreement between ordered vectors of features (ie. distinct genomic experiments). The package includes Correspondence-At-the-TOP (CAT) analysis.

Description

This function plots corresponding at the top (CAT) curves using overlap proportions computed by computeCat. A number of arguments can be used for a pretty display, and for annotating the plot, and adding the legend

Usage

plotCat(catData, whichToPlot = 1:length(catData),
     preComputedPI, size=500, main="CAT-plot",
     minYlim=0, maxYlim=1, col, pch, lty, cex=1, lwd=1,
     spacePts=10, cexPts=1, legend=TRUE, legendText,
     where="center", legCex=1,
     plotLayout=layout(matrix(1:2, ncol = 2, byrow = TRUE), widths = c(0.7, 0.3)), ...) 

Arguments

catData	The ouput list obtained from computeCat, containing the overlapping proportions among pairs of ordered vectors. Names in catData are used for annotating the legend if legendText is not provided (see below).
whichToPlot	numeric vector. Indexes corresponding to the elements of catData to be selected for displaying in the plot.
preComputedPI	numeric matrix. Probability intervals computed using the calcHypPI function. It is used to add grey shades to the plot corresponding to overlapping proportion probabilities based on the hypegeomtric distribution. If missing no PI will be added to the plot.
size	numeric. The number of top ranking features to be displayed in the plot.
main	character. The title of the plot, if not provided, main default is "CAT-plot".
minYlim	numeric. The lower numeric value of the y axis, to be displayed in the plot.
maxYlim	numeric. The upper numeric value of the y axis, to be displayed in the plot.
col	character or numeric. Vector specifying colors for CAT curves plotting. col default uses rainbow function to generate a color vector for all CAT curves in catData. When provided by the user, it will be recycled if needed.
pch	graphical parameter. pch specifies point types for annotating the CAT curves. If not provided, pch is created by default, and recycled if needed. See par for details.
lty	graphical parameter. The type of line for the plot. If not provided generated by default, recycled id needed. See par if needed.
cex	numeric. Standard graphical parameter useful for controlling axes and title annotation size. See par.
lwd	numeric. Standard graphical parameter useful for controlling line size. See par.
spacePts	numeric. Specifies the interval to be used for adding point labels on the CAT curves (evenly spaced along the x axis dimention).
cexPts	numeric. Graphical parameter useful for controlling points size used for annotating CAT-plot lines.
legend	logical. Wheter a legend should be added to the plot.
legendText	character. A vector used for legend creation. legendText default correspond to catData names.
where	character. The position of the plot where the legend will be created; where default is 'center', see legend.
legCex	numeric. Graphical parameter setting the font size for the legend text.
plotLayout	A layout matrix to arrange the plot and the legend. For further details see layout.
...	Other graphical parameters, currently passed only to legend (e.g. the number of columns to be used in the legend, or the legend background).

Details

This function uses outputs from computeCat and calcHypPI to plot the CAT curves and add grey shades corresponding to probability intervals. The default plot uses a pre-specified layout with separate areas for the plot and the legend. If not specified by the user, different points, colors and line types are used for the different CAT curves. If the CAT curves where computed using equal ranks (e.g. "equalRank" was passed to the method argument of the computeCat function), the user has the option of adding probability intervals to the plot. Such intervals must be pre-computed using the calcHypPI function.

Value

Produces an annotated CAT plot.

Note

In order to make the "best looking" plot for your needs you must play around with graphical parameters

Author(s)

Luigi Marchionni [e-mail]

References

Irizarry, R. A.; Warren, D.; Spencer, F.; Kim, I. F.; Biswal, S.; Frank, B. C.; Gabrielson, E.; Garcia, J. G. N.; Geoghegan, J.; Germino, G.; Griffin, C.; Hilmer, S. C.; Hoffman, E.; Jedlicka, A. E.; Kawasaki, E.; Martinez-Murillo, F.; Morsberger, L.; Lee, H.; Petersen, D.; Quackenbush, J.; Scott, A.; Wilson, M.; Yang, Y.; Ye, S. Q. and Yu, W. Multiple-laboratory comparison of microarray platforms. Nat Methods, 2005, 2, 345-350

Ross, A. E.; Marchionni, L.; Vuica-Ross, M.; Cheadle, C.; Fan, J.; Berman, D. M.; and Schaeffer E. M. Gene Expression Pathways of High Grade Localized Prostate Cancer. Prostate, 2011, 71, 1568-1578

Benassi, B.; Flavin, R.; Marchionni, L.; Zanata, S.; Pan, Y.; Chowdhury, D.; Marani, M.; Strano, S.; Muti, P.; and Blandino, G. c-Myc is activated via USP2a-mediated modulation of microRNAs in prostate cancer. Cancer Discovery, 2012, March, 2, 236-247

See Also

See computeCat, calcHypPI, rainbow, par, legend, and layout.

Examples

###load data
data(matchBoxExpression)

###the column name for the identifiers and  the ranking statistics
idCol <- "SYMBOL"
byCol <- "t"

####filter the redundant features using SYMBOL and t-statistics
matchBoxExpression <- lapply(matchBoxExpression, filterRedundant, idCol=idCol, byCol=byCol)

###select and merge into a matrix
mat <- mergeData(matchBoxExpression, idCol=idCol, byCol=byCol)

###COMPUTE CAT
cpH2L <- computeCat(mat, idCol=1, size=round(nrow(mat)/1),
	decreasing=TRUE, method="equalRank")

###CATplot without probability intervals
par(mar=c(3,3,2,1))
plotCat(cpH2L, main="CAT-plot, decreasing t-statistics",
	cex=1, lwd=2, cexPts=1.5, spacePts=15,
	legend=TRUE, where="center",
	legCex=1, ncol=1)

###compute probability intervals
confInt <- calcHypPI(data=mat)

###CATplot with probability intervals
par(mar=c(3,3,2,1))
plotCat(cpH2L, main="CAT-plot, decreasing t-statistics, probability intevals",
	cex=1, lwd=2, cexPts=1.5, spacePts=15,
	legend=TRUE, where="center",
	legCex=1, ncol=1) 

marchion/matchBox documentation built on May 9, 2019, 4:07 p.m.