simulate: Simulate an object.

In ybukhman/SynergyScreen: Analyze and simulate synergy screen data

Description

Depending on the input object, randomly set ic50 and m parameters of median effect equations and simulate dose-response data for a single experiment or an entire screen

Usage

simulate(object, ...)

## S4 method for signature 'Compound'
simulate(object, log.ic50.cv = 1, m.range = c(0.5,
  2.5), replace = F)

## S4 method for signature 'DRE'
simulate(object, compound, control.response = 1,
  log.ic50.cv = 1, m.range = c(0.5, 2.5), rel.resp.noise = 0.1,
  replace = F)

## S4 method for signature 'SynergyScreen'
simulate(object, control.response = 1,
  plate.bias.sd = 0.2, blank.response = 0.05, rel.resp.noise = 0.1,
  log.ic50.cv = 1, m.range = c(0.5, 2.5), digits = 3, replace = F)

Arguments

object	object of a supported class, e.g. Compound, DRE or SynergyScreen
...	method-specific arguments: see below
log.ic50.cv	coefficient of variation, i.e sd/mean, of the distribution that log(IC50) is selected from, numeric
m.range	the range of values to select m from, numeric vector of length 2
replace	replace pre-existing values of @ic50 and @m? logic
compound	list of Compound objects (DRE method)
control.response	"true" response value (e.g. OD) of untreated control, numeric
rel.resp.noise	average relative noise of the response variable, i.e. mean(random.error/response)
plate.bias.sd	standard deviation of plate bias values, see Details, numeric
blank.response	average response value of blank wells, numeric
digits	number of decimal places in dose values (SynergyScreen method)

Details

ic50 and m are parameters of the median effect equations for compounds and mixtures. They are simulated as follows:

1. log(ic50) of a compound is chosen from a normal distribution centered around 0.5*(log([e-mail]) + log([e-mail])) with coefficient of variation specified by parameter log.ic50.cv

2. IC50 of a mixture is chosen from a log-normal distribution centered around 1/(sum(fraction[i]/compound[i]@ic50)) with coefficient of variation specified by parameter log.ic50.cv

3. m is chosen from a uniform distribution between m.range[1] and m.range[2]

If Compound object(s) in the input already have their @ic50 and @m slots populated and replace is set to FALSE, the pre-existing values are retained.

For dose-response experiments, represented by DRE objects, response at each dose is chosen from a normal distribution with mean determined by the median response equation, i.e. expected.response = control.response / ((dose/ic50)^m + 1), and standard.deviation = expected.response*rel.resp.noise

In a multi-plate screen, plate bias is multiplicative. All wells that contain cell cultures are multiplied by a bias value. The bias values are chosen from a normal distribution with the mean of 1 and standard deviation of plate.bias.sd

Value

object of the same class as the input object

Methods (by class)

• Compound: Simulate Compound dose-response characteristics

  Randomly set ic50 and m parameter of the median effect equation for compound object

• DRE: Simulate DRE data

  Simulate dose-response experiment data based on the median effect equation with randomly set parameters IC50 and m

• SynergyScreen: Simulate a synergy screening experiment

  Simulate an entire screen, including plate biases, control wells, median effect equation parameters for compounds and mixtures, and dose-response experiment data

See Also

SynergyScreen, Compound, DRE

Examples

#  Load a set of compounds, initialize a SynergyScreen object, and generate a design
cpds3 = readCompoundFile(system.file("extdata/15_cpds_simulation/compounds_3.csv",
                                     package="SynergyScreen"))
screen3 = new("SynergyScreen",compound_list=cpds3)
screen3 = generateDesign(screen3)

#  Simulate dose-response characteristics of compounds, mixtures and dose-response data
screen3_sim = simulate(screen3)

#  Simulation sets ic50 and m
compound(screen3,"Cpd1")
compound(screen3_sim,"Cpd1")
true.ic50(dre(screen3,"Cpd1"))
true.ic50(dre(screen3_sim,"Cpd1"))

#  Simulation sets response values
response(dre(screen3,"Cpd1"))
response(dre(screen3_sim,"Cpd1"))

ybukhman/SynergyScreen documentation built on May 4, 2019, 2:31 p.m.