getFitDT: The optimization function that finds the scaling factor for...

In saezlab/CNORdt: Add-on to CellNOptR: Discretized time treatments

Description

This function is called from gaBinaryDT. Using the model passed as input, it finds a scaling factor that minimizes the mean squared error between the data from the boolean simulation and the experimental data. A spline is fitted to the experimental data to allow this.

Usage

getFitDT(simResults, CNOlist, model, indexList, sizeFac = 1e-04, NAFac = 1, nInTot, boolUpdates, lowerB, upperB)

Arguments

simResults	The simulation results as output from simulatorDT
CNOlist	A CNOlist on which the score is based (based on all valueSignals).
model	A model list.
indexList	A list of indexes of species stimulated/inhibited/signals, as produced by indexfinder applied on the model and CNOlist above.
sizeFac	The scaling factor for the size term in the objective function, default to 0.0001.
NAFac	The scaling factor for the NA term in the objective function, default to 1.
nInTot	The number of inputs in the model prior to cutting, used to normalise the size penalty.
boolUpdates	The number of synchronous updates performed by the boolean simulator.
lowerB	The lower bound for the optimized value of the scaling factor.
upperB	The upper bound for the optimized value of the scaling factor.

Details

The function optim() is used to find the optimal scaling factor.

Value

This function returns a list with elements:

score	The mean squared error between simulation and experiment with NA and model size penalties.
estimate	The scaling factor used to compare boolean and real data.
xCoords	The x-axis coordinates after multiplication with the scaling factor.
yInter	The interpolated values of the experimental data.
yBool	The boolean simulation results at each time point.

Author(s)

A. MacNamara

See Also

gaBinaryDT, simulatorDT

Examples

# this function is usually contained within gaBinaryDT
# but the output can be viewed as follows:

library(CellNOptR)
library(CNORdt)
data(CNOlistPB, package="CNORdt")
data(modelPB, package="CNORdt")

# pre-processing
indexOrig <- indexFinder(CNOlist=CNOlistPB, model=modelPB, verbose=TRUE)
fields4Sim <- prep4sim(model=modelPB)

boolUpdates = 10
simResults <- simulatorDT(
  CNOlist=CNOlistPB,
  model=modelPB,
  simList=fields4Sim,
  indices=indexOrig,
  boolUpdates=boolUpdates
)
simResults = convert2array(simResults, dim(CNOlistPB$valueSignals[[1]])[1],
length(modelPB$namesSpecies), boolUpdates)

optimRes <- getFitDT(
  simResults=simResults,
  CNOlist=CNOlistPB,
  model=modelPB,
  indexList=indexOrig,
  boolUpdates=boolUpdates,
  lowerB=0.8,
  upperB=10,
  nInTot=length(which(modelPB$interMat == -1))
)

saezlab/CNORdt documentation built on May 17, 2019, 1:59 p.m.