fitOneExp: Fit multiple dose response models for a single drug in a...

In lshen1/drexplorer2: Dose-response Explorer2

Description

The dose response mode is usually fitted with original dose (not log10 transformed). The computed ICx values however, can be in either log10 scale or the original scale.

Usage

fitOneExp(
  dat,
  drug = NA,
  cellLine = NA,
  unit = NA,
  models = c("LL.3", "LL.3u", "sigEmax", "logistic", "linlog"),
  cols = NA,
  log.d = TRUE,
  percentile = seq(0.1, 0.9, by = 0.1),
  alpha = 0.01,
  fitCtr = FALSE,
  standardize = TRUE,
  interpolation = TRUE,
  plot = FALSE,
  transparency = 1,
  ...
)

Arguments

dat	a 2-column data frame with first column as dose and second column response. Controls are decided by dose = 0.
drug	drug for this analysis.
cellLine	cell line for this analysis.
unit	unit of drug concentration.
models	models the user may specify.
cols	colors of dose-response curves.
log.d	in computed ICx values, whether to return the log10 dose or the dose without this transformation.
percentile	IC percentile.
alpha	outlier significance level.
fitCtr	whether the model is fitted with control data.
standardize	whether to standardize (scale) the data based on control points. This should be disabled when no control data is supplied.
interpolation	whether calculate ICx through interpolation or through root finding. This parameter is passed to computeIC() function. When interpolation = TRUE, ICx value will be bounded by observed minimum and maximum dose; otherwise, ICx is calculated through root finding and thus can be outside this boundary (extrapolated).
plot	whether to draw the dose response curve.
transparency	a value between 0 and 1 specifying transparency through alpha blending; transparency = 0 means totally transparent.
...	additional parameters to plotOneExp().

Value

a list containing elements with:

fits,

models,

cols,

ICmat, IC matrix from all specified models as well as RSE and model name

ICx, IC values from the best model

datWithOutlierStatus the input data with outlier status appended

bestModel the best model by RSE

RSEs

Examples

fitExp <- fitOneExp(ryegrass[, c(2, 1)], drug = '', cellLine = '', unit = '',
                    models = c('sigEmax', 'LL.4', 'LL.5', 'LL.3', 'LL.3u', 'logistic'),
                    alpha = 0.05, interpolation = TRUE) 

lshen1/drexplorer2 documentation built on June 2, 2020, 9:27 p.m.