responseMap {braidReports}

Visualizing Combination Response Maps

Description

Produces a ggplot2 plot depicting some response variable as a function of two input variables; generally this will be some biological effect as a function of two compound concentrations.

Usage

## Default S3 method:
responseMap(model,data,irreg=FALSE,interpolate=TRUE,margins=TRUE,
			logscale=TRUE,zlims=NULL,raster=TRUE)
## S3 method for class 'formula'
responseMap(model,data,...)

Arguments

model

a two-column array containing the values of input 1 and input 2 in each data point, or a symbolic formula (e.g. act ~ conc1+conc2) specifying which variables are to be visualized

data

if model is an array, a vector of values in response to inputs 1 and 2; if model is a formula, a data frame containing the columns specified in formula

irreg

reflects whether or not data points lie in a regularly-spaced grid in the space of data points to be plotted

interpolate

determines whether the plot will produce a smoothed response surface using Gaussian interpolation, or if the space will be visualized as a set of discrete measurements using nearest neighbor segmentation

margins

determines whether input pairs in which one of the two inputs is zero will be plotted; applies only to logarithmically scaled plots

logscale

determines whether the input variables will be plotted on a logarithmic scale, or a standard linear scale. Because this function is intended primarily for visualizing combined action dose-response, a logarithmic dose-pair space is the default.

zlims

an optional parameter specifying lower and upper bounds on the response variable. If present, all effect values outside the given range will be clipped to the minimum and maximum values

raster

by default, interpolated response surfaces will be rendered using the ggplot2 function geom_raster, a faster and optimized raster image version of geom_tile. In certain settings, however, geom_raster can produce unusual appearance. Setting this parameter to FALSE will force the plot to be rendered using geom_tile

...

not used

Details

This function produces and returns a ggplot2 plot, but does not draw it to a graphics device. As the strength of ggplot2 is in the ability to annotate, build on, and style plots using additional layers and themes, this function adds no theme information, and very little scale information. To produce a more aesthetic plot, it is recommended that plot be given a custom color scale, axis labels, and theme by using the appropriate ggplot2 functions. For convenience, the functions formatResponseMap and formatDifferenceMap format and add labels to plots produced by responseMap, but any ggplot2 changes that are desired can be made. If you wish to add additional layers (such as contour curves), the data variables available for mapping aesthetics are x, y, and z.

Value

A ggplot2 plot object, depicting the response variable (represented by the plot\'s \'fill\' variable) as a function of two inputs (the x- and y-dimensions).

Author(s)

Nathaniel R. Twarog

References

Twarog, N.R., Stewart, E., Vowell Hamill, C., and Shelat, A. BRAID: A Unifying Paradigm for the Analysis of Combined Drug Action. Scientific Reports In Press (2016).

See Also

formatResponseMap, formatDifferenceMap, voronoi

Examples

data(es8olatmz,package="braidrm")
rmap <- responseMap(act~conc1+conc2,es8olatmz)
formatResponseMap(rmap,xl="Olaparib",yl="TMZ",zl="Survival")
rmap <- responseMap(act~conc1+conc2,es8olatmz,interpolate=FALSE,margins=FALSE)
formatDifferenceMap(rmap,zcenter=-1,xl="Olaparib",yl="TMZ",zl="Survival")