metaboGSE (version 1.2.5)

fitness: Fitness of gene removal-based submodels with different gene rankings

Description

This function computes the fitness of submodels by removing genes in different gene rankings.

Usage

fitness(
  model,
  ranks,
  rescue.weight = NULL,
  step = 1,
  draw.num = 0,
  obj.react = NA,
  mc.cores = 1,
  timeout = 12,
  tol = SYBIL_SETTINGS("TOLERANCE"),
  solver = SYBIL_SETTINGS("SOLVER"),
  method = SYBIL_SETTINGS("METHOD")
)

Arguments

model

An object of class modelorg indicating the weighted rescue model obtained from the rescue process.

ranks

A list of data frames of scores for ranking genes, with gene per row, e.g. data.frame(pkm=pkm expression, rel=relative expression).

rescue.weight

A vector of rescue reaction weights. Default: NULL, the weights are computed from the given model with gene.num=1.

step

An integer indicating the step in numbers of genes to remove. Default: 1, gene-by-gene removal. When there are many genes in the model, the step is multiplied by an exponent of 2 for later removals. This is to reduce the computing time for non-informative sub-models at the end of the series.

draw.num

Number of random draws. Default: 0.

obj.react

A string indicating objective reaction ID. Default: reaction producing BIOMASS.

mc.cores

The number of cores to use (at least 1), i.e. at most how many child processes will be run simultaneously. Default: 1.

timeout

The maximum time in seconds to allow for LP call to return. Default: 12.

tol

The maximum value to be considered null. Default: SYBIL_SETTINGS("TOLERANCE").

solver

sybil solver. Default: SYBIL_SETTINGS("SOLVER").

method

sybil method. Default: SYBIL_SETTINGS("METHOD").

Value

An object of class scoreGeneDel for the submodel construction simulation.

Examples

# NOT RUN {
data(Ec_core)
mod <- rescue(Ec_core, target=0.1)
mod.weight <- changeObjFunc(mod$rescue, react=rownames(mod$coef), obj_coef=mod$coef)
ranks <- list(
   rep.1=data.frame(
       expr=setNames(rnorm(length(sybil::allGenes(mod.weight)), mean=5, sd=4),
           sybil::allGenes(mod.weight))),
   rep.2=data.frame(
       expr=setNames(rnorm(length(sybil::allGenes(mod.weight)), mean=5, sd=4.1),
           sybil::allGenes(mod.weight))))
fn <- fitness(model=mod.weight, ranks=ranks, step=200, draw.num=1)
# }