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

Description Usage Arguments Note Author(s) See Also Examples

Description

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

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
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 [email protected], Ning Leng [email protected], and previous team members (see DESCRIPTION)

See Also

SummaryTwoGroups

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
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.