sracipeSimulate: Simulate a gene regulatory circuit

In lusystemsbio/sRACIPE: Systems biology tool to simulate gene regulatory circuits

Simulate a gene regulatory circuit

Description

Simulate a gene regulatory circuit using its topology as the only input. It will generate an ensemble of random models.

Usage

sracipeSimulate(
  circuit = "inputs/test.tpo",
  config = config,
  anneal = FALSE,
  knockOut = NA_character_,
  numModels = 2000,
  paramRange = 100,
  prodRateMin = 1,
  prodRateMax = 100,
  degRateMin = 0.1,
  degRateMax = 1,
  foldChangeMin = 1,
  foldChangeMax = 100,
  hillCoeffMin = 1L,
  hillCoeffMax = 6L,
  integrateStepSize = 0.02,
  simulationTime = 50,
  nIC = 1L,
  nNoise = 0L,
  simDet = TRUE,
  initialNoise = 50,
  noiseScalingFactor = 0.5,
  shotNoise = 0,
  ouNoise_t = 1,
  scaledNoise = FALSE,
  outputPrecision = 12L,
  printStart = 50,
  printInterval = 10,
  stepper = "RK4",
  thresholdModels = 5000,
  plots = FALSE,
  plotToFile = FALSE,
  genIC = TRUE,
  genParams = TRUE,
  integrate = TRUE,
  rkTolerance = 0.01,
  timeSeries = FALSE,
  signalRate = 10,
  uniqueDigits = 4,
  convergThresh = 1e-12,
  numStepsConverge = 500,
  numConvergenceIter = 25,
  limitcycles = FALSE,
  LCSimTime = 10,
  LCSimStepSize = 0.01,
  maxLCs = 10,
  LCIter = 20,
  maxPeriods = 100,
  numSampledPeriods = 3,
  allowedPeriodError = 3,
  samePointProximity = 0.1,
  LCStepper = "RK4",
  paramSignalVals = data.frame(),
  geneClamping = data.frame(),
  nCores = 1L,
  ...
)

Arguments

circuit	data.frame or character. The file containing the circuit or
config	(optional) List. It contains simulation parameters like integration method (stepper) and other lists or vectors like simParams, stochParams, hyperParams, options, thresholds etc. The list simParams contains values for parameters like the number of models (numModels), simulation time (simulationTime), step size for simulations (integrateStepSize), when to start recording the gene expressions (printStart), time interval between recordings (printInterval), number of initial conditions (nIC), output precision (outputPrecision), tolerance for adaptive runge kutta method (rkTolerance), parametric variation (paramRange). The list stochParams contains the parameters for stochastic simulations like the number of noise levels to be simulated (nNoise), the ratio of subsequent noise levels (noiseScalingFactor), maximum noise (initialNoise), whether to use same noise for all genes or to scale it as per the median expression of the genes (scaledNoise), ratio of shot noise to additive noise (shotNoise). The list hyperParams contains the parameters like the minimum and maximum production and degration of the genes, fold change, hill coefficient etc. The list options includes logical values like annealing (anneal), scaling of noise (scaledNoise), generation of new initial conditions (genIC), parameters (genParams) and whether to integrate or not (integrate). The user modifiable simulation options can be specified as other arguments. This list should be used if one wants to modify many settings for multiple simulations.
anneal	(optional) Logical. Default FALSE. Whether to use annealing for stochastic simulations. If TRUE, the gene expressions at higher noise are used as initial conditions for simulations at lower noise.
knockOut	(optional) List of character or vector of characters. Simulation after knocking out one or more genes. To knock out all the genes in the circuit, use knockOut = "all". If it is a vector, then all the genes in the vector will be knocked out simultaneously.
numModels	(optional) Integer. Default 2000. Number of random models to be simulated.
paramRange	(optional) numeric (0-100). Default 100. The relative range of parameters (production rate, degradation rate, fold change).
prodRateMin	(optional) numeric. Default 1. Minimum production rate.
prodRateMax	(optional) numeric. Default 100. Maximum production rate.
degRateMin	(optional) numeric. Default 0.1. Minimum degradation rate.
degRateMax	(optional) numeric. Default 1. Maximum degradation rate.
foldChangeMin	(optional) numeric. Default 1. Minimum fold change for interactions.
foldChangeMax	(optional) numeric. Default 100. Maximum fold change for interactions.
hillCoeffMin	(optional) integer. Default 1. Minimum hill coefficient.
hillCoeffMax	(optional) integer. Default 6. Maximum hill coefficient.
integrateStepSize	(optional) numeric. Default 0.02. step size for integration using "EM" and "RK4" steppers.
simulationTime	(optional) numeric. Default 50. Total simulation time. Only used for stochastic and time series simulations. For adjusting deterministic simulation run times, see numConvergenceIter.
nIC	(optional) integer. Default 1. Number of initial conditions to be simulated for each model.
nNoise	(optional) integer. Default 0. Number of noise levels at which simulations are to be done. Use nNoise = 1 if simulations are to be carried out at a specific noise. If nNoise > 0, simulations will be carried out at nNoise levels as well as for zero noise. "EM" stepper will be used for simulations and any argument for stepper will be ignoired.
simDet	(optional) logical. Default TRUE. Whether to simulate at zero noise as well also when using nNoise > 0.
initialNoise	(optional) numeric. Default 50/sqrt(number of genes in the circuit). The initial value of noise for simulations. The noise value will decrease by a factor noiseScalingFactor at subsequent noise levels.
noiseScalingFactor	(optional) numeric (0-1) Default 0.5. The factor by which noise will be decreased when nNoise > 1.
shotNoise	(optional) numeric. Default 0. The ratio of shot noise to additive noise.
ouNoise_t	(optional) numeric. Default 1. Correlation time parameter for OU noise. Only used when "EM_OU" stepper is selected
scaledNoise	(optional) logical. Default FALSE. Whether to scale the noise in each gene by its expected median expression across all models. If TRUE the noise in each gene will be proportional to its expression levels.
outputPrecision	(optional) integer. Default 12. The decimal point precison of the output.
printStart	(optional) numeric (0-simulationTime). Default simulationTime. To be used only when timeSeries is TRUE. The time from which the output should be recorded. Useful for time series analysis and studying the dynamics of a model for a particular initial condition.
printInterval	(optional) numeric (integrateStepSize- simulationTime - printStart). Default 10. The separation between two recorded time points for a given trajectory. To be used only when timeSeries is TRUE.
stepper	(optional) Character. Stepper to be used for integrating the differential equations. The options include "EM" for Euler-Maruyama O(1), "RK4" for fourth order Runge-Kutta O(4) and "DP" for adaptive stepper based Dormand-Prince algorithm. The default method is "RK4" for deterministic simulations and the method defaults to "EM" for stochastic simulations.
thresholdModels	(optional) integer. Default 5000. The number of models to be used for calculating the thresholds for genes.
plots	(optional) logical Default FALSE. Whether to plot the simuated data.
plotToFile	(optional) Default FALSE. Whether to save the plots to a file.
genIC	(optional) logical. Default TRUE. Whether to generate the initial conditions. If FALSE, the initial conditions must be supplied as a dataframe to circuit$ic.
genParams	(optional) logical. Default TRUE. Whether to generate the parameters. If FALSE, the parameters must be supplied as a dataframe to circuit$params.
integrate	(optional) logical. Default TRUE. Whether to integrate the differential equations or not. If FALSE, the function will only generate the parameters and initial conditions. This can be used iteratively as one can fist generate the parameters and initial conditions and then modify these before using these modified values for integration. For example, this can be used to knockOut genes by changing the production rate and initial condition to zero.
rkTolerance	(optional) numeric. Default 0.01. Error tolerance for adaptive integration method.
timeSeries	(optional) logical. Default FALSE. Whether to generate time series for a single model instead of performing RACIPE simulations.
signalRate	(optional) numeric. Default 10. The factor by which the differential equations for fast genes are multiplied by. Fast genes only have inward signaling interactions of types 5 or 6
uniqueDigits	(optional) integer. Default 4. Deterministic simulations with nIC > 1 count the number of unique steady states per model by truncating the expression by this many digits and counting the number of unique expressions
convergThresh	(optional) numeric. Default 1e-12. The threshold for convergence to a steady state for deterministic simulations.
numConvergenceIter	(optional) integer. Default 25. The total number of convergence test iterations to run per model initial condition in deterministic simulations.
limitcycles	(optional) logical. Default FALSE. Whether to check for limit cycles in deterministic simulations.
LCSimTime	(optional) numeric. Default 10. The length of each iteration for the secondary limit cycle simulation
LCSimStepSize	(optional) numeric. Default 0.01. The integration step size for the secondary limit cycle simulation.
maxLCs	(optional) integer. Default 10. The maximum allowable number of limit cycles that can be detected for each model.
LCIter	(optional) integer. Default 20. The number of iterations to run in the secondary limit cycle simulation
maxPeriods	(optional) integer. Default 100. The number of periods to count in the distance function constructed by the limit cycle detection algorithm
numSampledPeriods	(optional) integer. Default 3. The number of times the limit cycle detection algorithm tries to calculate the period of the simulated trajectory using the local minima in a constructed distance function.
allowedPeriodError	(optional) integer. Default 3. The allowed difference in the sampled periods for a trajectory from the limit cycle detection algorithm. Decrease this value to make for more stringent limit cycle detection
samePointProximity	(optional) numeric. Default 0.1. The max allowed square euclidean difference between two minima of the distance function constructed by the limit cycle detection algorithm before they are considered different points in phase space.
LCStepper	(optional) Character. Default "RK4" The integration method used for the limit cycle simulation. The options include "E" for first order Euler and "RK4" for fourth order Runge Kutta. Any other input will cause the simulation to use the Euler method
paramSignalVals	(optional) Data Frame. Default data.frame(). The first column must be a vector of time values with the first element as 0 and the last element as simulationTime. The other column names must be valid production or degradation parameter names for the circuit. Use sracipeGenParamNames() to generate valid parameter names. The first column must be a vector of time values with the first element as 0 and the last element as simulationTime.
geneClamping	(optional) Data Frame. Default data.frame(). The column names must be genes in the circuit. The number of columns must either be one or equal to numModels. If the number of columns is one, the selected genes are clamped to those values for every model. Otherwise, the gene is clamped to the value of the corresponding row for a particular model.
nCores	(optional) integer. Default 1 Number of cores to be used for computation. Utilizes multisession from doFuture pacakge, as well and doRNG package.
...	Other arguments
numStepConverge	(optional) integer Default 500. The number of integration steps between convergence tests for deterministic simulations.

Value

RacipeSE object. RacipeSE class inherits SummarizedExperiment and contains the circuit, parameters, initial conditions, simulated gene expressions, and simulation configuration. These can be accessed using correponding getters.

Related Functions

sracipeSimulate, sracipeKnockDown, sracipeOverExp, sracipePlotData

Examples

data("demoCircuit")
rSet <- sRACIPE::sracipeSimulate(circuit = demoCircuit)

lusystemsbio/sRACIPE documentation built on Dec. 9, 2024, 8:25 a.m.