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

shieldio_badge <- function(name, text, color, hover = NULL, href = NULL, ..., 
  url = 'https://img.shields.io/badge/{{name}}-{{text}}-{{color}}.svg',
  style = 'flat-square') {

  .dots <- list(...)
  name  <- RCurl::curlEscape(whisker::whisker.render(name, .dots))
  text  <- RCurl::curlEscape(whisker::whisker.render(text, .dots))
  color <- whisker::whisker.render(color, .dots)

  if (is.numeric(c <- type.convert(color)))
    color <- switch(letters[[floor(pmax(pmin(c, 1), 0) * 6 + 1)]],
      g = 'brightgreen',
      f = 'brightgreen',
      e = 'green',
      d = 'yellowgreen',
      c = 'yellow',
      b = 'orange',
      a = 'red')

  color <- RCurl::curlEscape(color)  
  url <- whisker::whisker.render(paste(url, paste0('style=', style), sep = '?'))
  hover <- if (is.null(hover)) name else hover
  whisker::whisker.render('[![{{hover}}]({{url}})]({{href}})')
}

failed_tests = tryCatch(devtools::test(), error = function(e) NULL)
failed_tests = if (!is.null(failed_tests) && 'failed' %in% names(as.data.frame(failed_tests))) sum(failed_tests$failed) else 1
build_badge <- shieldio_badge(
  'build', '{{p}}', '{{d}}', 'Build Status', 
  p = if (failed_tests == 0) 'passing' else 'failing', 
  d = if (failed_tests == 0) 'brightgreen' else 'red')

coverage = covr::percent_coverage(covr::package_coverage()); 
coverage_badge <- shieldio_badge(
  'coverage', '{{p}}%', '{{d}}', 'Code Coverage', 
  p = round(coverage,1), 
  d = (coverage/100-0.8)/0.2)

gClinBiomarker

r build_badge r coverage_badge

Overview

gClinBiomarker is an R package that contains functions to perform baseline and longitutinal biomarker analyses

Source code: https://github.roche.com/Rpackages/gClinBiomarker

Project page: https://pages.github.roche.com/Rpackages/gClinBiomarker/

Documentations could be found under the "article" tab of the project page. Two documents are provided: - A user vignette demonstrates how gClinBiomarker package may help in your biomarker analysis - An exapmle use case document which contains more detailed use cases

pdf version of these two documents can be found here.

gClinBiomarker also provide a set of Rmarkdown templates that allow you to generate biomarker analysis report by "one click". The R markdown templates can be found at here, along with some slide decks.

Getting Started

Installation

To install this package from R, use install_github() function from the devtools package

In R, type:

## install.packages("devtools")
library(devtools)
install_github("RPackages/gClinBiomarker", host="https://github.roche.com/api/v3")

Note that on bce (r.bas.roche.com), the default R is from an older version. To install, type

install_github("RPackages/gClinBiomarker", host="github.roche.com/api/v3")

Getting the Documentation

Use the command vignette(package = 'gClinBiomarker') to view a list of avaialble vignettes for the gClinBiomarker package. Then use the command vignette(<vignett_name>) to view the relevant documentation.

gClinBiomarker Analysis and Workflows

Supported endpoint and biomarker types

Endpoints

• Time-to-event
• Binary
• Continuous
• Longitudinal

Biomarkers

• Continuous
• Ordinal
• Categorical
• Longitudinal

Analysis

• 2-arm study: predictive and prognostic biomarker analysis
• 1-arm study: prognostic biomarker analysis

Workflow

Step 1:

Is your biomarker population your full patient population? Take a look at these functions:

SummaryVars(), CompareKM(), PlotRspBar()

Step 2:

Is your biomarker response a dynamic range? Does it have a skewed distribution? Is it correlated with clinical variables?

PlotProperty(), PlotTabForestMulti()

Step 3:

Is your biomarker response associated with a clinical outcome? Is it prognostic or predictive? Is there an optimal biomarker cutoff with a consistent trend?

PlotTabForestBiomarker(), PlotSTEPP()

Step 4:

Estimate the clinical benefit within a biomarker subgroup.

PlotKM(), PlotLong(), PlotRspBar(), CoxTab(), LogRankTab()

# ![screen shot 2017-09-28 at 1 44 01 pm](https://media.github.roche.com/user/48/files/431d398a-a453-11e7-8801-1c6915156185)
# ![screen shot 2017-09-28 at 1 44 13 pm](https://media.github.roche.com/user/48/files/47cad168-a453-11e7-85f7-deee2f7604ab)
# ![screen shot 2017-09-28 at 1 44 24 pm](https://media.github.roche.com/user/48/files/4ad58466-a453-11e7-80be-9af0c23fcedd)
# ![screen shot 2017-09-28 at 1 44 34 pm](https://media.github.roche.com/user/48/files/507dd1a2-a453-11e7-8216-442f15bd500a)
# ![screen shot 2017-09-28 at 1 44 43 pm](https://media.github.roche.com/user/48/files/53ab85e0-a453-11e7-825d-718f28d1fca3)

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