linearQuadraticModel: Fit linear-quadratic curves to dose-response data

In bhklab/RadioGx: Analysis of Large-Scale Radio-Genomic Data

Fit linear-quadratic curves to dose-response data

Description

This function fits a linear-quadratic curve to dose-response data.

Usage

linearQuadraticModel(
  D,
  SF,
  lower_bounds = c(0, 0),
  upper_bounds = c(1, 1),
  scale = 5,
  family = c("normal", "Cauchy"),
  median_n = 1,
  trunc = FALSE,
  verbose = FALSE
)

Arguments

D	vector of radiation doses
SF	vector of survival fractions corresponding to the doses
lower_bounds	vector of length 2 containing minimum allowed values of fitted alpha and beta, respectively
upper_bounds	vector of length 2 containing maximum allowed values of fitted alpha and beta, respectively
scale	parameter of the assumed error distribution of the data; see sdetails
family	family of distributions of the error terms in the data; currently supported options are "normal" and "cauchy"
median_n	see details
trunc	should survival fractions be truncated downward to 1? Defaults to FALSE.
verbose	'verbose' outputs warnings that are otherwised suppressed when the function sanity-checks user inputs. 'median_n' denotes the number of distributions from family 'family' that are medianned. (Note that setting n = 1 (the default) is equivalent to using a simple normal or cauchy distribution without taking any medians.)

Value

numeric The estimated alpha and beta values

Examples

linearQuadraticModel(c(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10),
 c(1.1, 0.8, 0.7, 0.45, 0.15, -0.1, -0.1, -0.4, -0.65, -0.75, -1.1))

bhklab/RadioGx documentation built on Oct. 6, 2023, 8:27 a.m.