calculateResponseScore: Calculates a score based on principle component analysis

In smm52/timeBlocks: Converts Serial Measurement Data into Longitudinal Trajectory of Equally-sized Time Blocks

Calculates a score based on principle component analysis

Description

The function calcuculates a score for biomarker data represented as time blocks. The score is calculated via PCA.

Usage

calculateResponseScore(
  responseDf,
  idColumn = "PATIENT",
  blockColumn = "block",
  sexColumn = NA,
  responseColumns,
  confounderColumns = NA
)

Arguments

responseDf	the data frame that contains the time blocks and the biomarker data per time block
idColumn	name of ID column
blockColumn	name of time block column
sexColumn	name of column that contains sex information (optional). If provided, the score will be adjusted for sex.
confounderColumns	names of columns that contain confounders (optional). If provided, the score will be adjusted for these confounders using linear regression
responseColumn	names of the columns containing the patient block value for a biomarker (e.g. vector of string names). The score will be based only on these columns.

Value

the response data frame with an added column phenoScore that shows the computed score

Examples

## Not run: 
results <- read_tsv('../data/responseSpreadsheet.txt')

resultsCor <- calculateResponseScore(results,responseColumns = c('sbp','dbp','aerAll','acrAll','egfr'), sexColumn = 'SEX')

## End(Not run)

smm52/timeBlocks documentation built on Nov. 27, 2022, 3:54 a.m.