PlotSTEPP: Create a STEPP (Subpopulation Treatment Effect Pattern Plot)

In lengning/gClinBiomarker: gClinBiomarker: an R package for clinical biomarker analyses

Description

This function creates a STEPP from the given point estimates and confidence intervals at desired percentiles.

Usage

PlotSTEPP(data, outcome.var, outcome.class, trt = NULL, var,
  covariate = NULL, strata = NULL, placebo.code = NULL,
  active.code = NULL, quantile.type = 1, alpha = 0.05,
  window.width = 0.25, min.pt = NULL, max.pt = NULL, by = 0.05,
  yrange.lower = NULL, yrange.upper = NULL, xticks = NULL,
  show.refline = TRUE, refline.color = "grey", show.refline.ac = TRUE,
  refline.color.ac = "lightblue", estimate.color = "blue",
  estimate.lty = 1, estimate.lwd = 2, surv.conf.type = "plain",
  ties = "efron", ci.color = "black", ci.lty = 2, ci.lwd = 1,
  ci.shade = TRUE,
  plot.title = "STEPP: Subgroup Treatment Effect Pattern Plot",
  sub.title = NULL, xlabel = "Biomarker Percentile", ylabel = NULL,
  show.legend = TRUE, legend.loc = "topright",
  legend.text = c("Point Estimate", paste((1 - alpha) * 100,
  " percent Confidence Interval", sep = "")), legend.col = c(estimate.color,
  ci.color), legend.lty = c(estimate.lty, ci.lty),
  legend.lwd = c(estimate.lwd, ci.lwd), legend.bty = "n", bm.digits = 2,
  actual.scale = FALSE, equal.in.LL = TRUE, equal.in.UL = FALSE,
  pdf.name = NULL, pdf.param = list(width = 11, height = 8.5),
  par.param = list(mar = c(4, 4, 3, 2)), csv.name = NULL)

Arguments

data	input data frame. Rows are patients and columns are variables (e.g. demographics variables, time to event variables, biomarker variables, treatment indicator, etc.). One patient per row.
outcome.var	outcome variable. In case of a 'survival', variable, it will be a vector of two variables: 1) time to event 2) censorship
outcome.class	outcome class of the 'outcome' variable. Can be either "continuous", "binary", or "survival".
trt	name of the treatment variable.
var	name of the biomarker variable.
covariate	vector specifying the covariate variables. This can be added to adjust for in the analysis for survival and continuous outcome variable classes.
strata	vector specifying the stratification variables. This can be added for the survival outcome variable class.
placebo.code	name of the control group within the treatment variable
active.code	name of the treatment/experimental group within the treatment variable
quantile.type	an integer between 1 and 9 selecting one of the nine quantile algorithms. See quantile. Default is 2.
alpha	confidence level (CI) for point estimate, i.e. 0.05 for 95 percent CI. Default is 0.05.
window.width	width of each window. Default is 0.25.
min.pt	minimum center.pt. Default is NULL.
max.pt	maximum center.pt. Default is NULL.
by	size of 'slide', i.e. speed of the window that moves along the x-axis. Default is 0.05.
yrange.lower	value of the lower y-axis. Default is NULL.
yrange.upper	value of the upper y-axis. Default is NULL.
xticks	x-tick marks. Default is NULL.
show.refline	if TRUE, the reference line will be displayed. Default is TRUE.
refline.color	color of the reference line. Default is "grey".
show.refline.ac	if TRUE, the reference line for effect in All Comers will be displayed. Default is TRUE.
refline.color.ac	color of the reference line for effect in All Comers. Default is "lightblue".
estimate.color	color of the estimate line. Default is "blue".
estimate.lty	type of the estimate line. Default is 1.
estimate.lwd	width of the estimate line. Default is 2.
surv.conf.type	confidence interval type. Default is "plain". see conf.type in survfit.
ties	Default is "efron". To match internal sas results, use "exact". See parameter "ties" in coxph.
ci.color	color of the CI lines. Default is "black".
ci.lty	type of the CI lines. Default is 2.
ci.lwd	width of the CI lines. Default is 1.
ci.shade	if TRUE, the area between the lower and upper CI lines will be shaded in gray. Default is TRUE.
plot.title	title of the plot. Default is "STEPP: Subgroup Treatment Effect Pattern Plot".
sub.title	subtitle of the plot, this will be displayed on the bottom of the plot. Default is NULL.
xlabel	x-axis label. Default is "Biomarker Percentile".
ylabel	y-axis label. Default is NULL.
show.legend	if TRUE, a legend will be displayed. Default is TRUE.
legend.loc	location of the legend to be displayed. Default is "topright".
legend.text	text to be displayed in the legend. Default is c("Point Estimate",paste((1-alpha)*100," percent Confidence Interval",sep=""))
legend.col	colors of the legend. Default is c(estimate.color, ci.color).
legend.lty	type of the legend lines. Default is c(estimate.lty, ci.lty).
legend.lwd	width of the legend lines. Default is c(estimate.lwd, ci.lwd).
legend.bty	type of box for the legend. Default is "n".
bm.digits	digits to be displayed/used for the lower and upper confidence level estimates. Default is 2.
actual.scale	if TRUE, it generates the figure using the actual scale instead of the scale in percentage. Default is FALSE.
equal.in.LL, equal.in.UL	if both are TRUE, window is defined using >= and <= (in legend: "[ ]"). if (TRUE, FALSE), window is defined using >= and < (in legend: "[ )"). if (FALSE, TRUE), window is defined using > and <= (in legend: "( ]"). if both are FALSE, window is defined using > and < (in legend: "()"). Default is (TRUE, FALSE).
pdf.name	name of output pdf file. If it's NULL (default), the plots will be displayed but not saved as pdf.
pdf.param	a list of parameters that define pdf graphics device. See pdf. Default is list(width=11, height=8.5).
par.param	a list of parameters that define graphcial parameters. See par. Default is list(mar=c(4,4,3,2)).
csv.name	csv file name (includes numbers used in the graphs). If NULL (default), the function will return the result.

Note

Patient data without corresponding biomarker data are automatically removed. For survival data, censorship variable is 1 if an event happened, 0 if censored.

Author(s)

Alexey Pronin [e-mail], Ning Leng [e-mail], and previous team members (see DESCRIPTION)

See Also

SummaryTwoGroups

Examples

data(input)
PlotSTEPP(data = input,
         outcome.var = c("PFS", "PFS.event"),
         outcome.class = "survival",
         trt = "Arm",
         var = "KRAS.exprs",
         covariate = "Sex",
         strata = "Age",
         placebo.code = "CTRL",
         active.code = "TRT")

lengning/gClinBiomarker documentation built on May 9, 2019, 2:55 p.m.