sysBiolAlg_room-class: Class '"sysBiolAlg_room"'

In SysBioChalmers/sybil: Efficient Constrained Based Modelling

Class "sysBiolAlg_room"

Description

The class sysBiolAlg_room holds an object of class optObj which is generated to meet the requirements of the ROOM algorithm.

Details

The initialize method has the following arguments:

model

An object of class modelorg.

wtflux

A numeric vector holding an optimal wild type flux distribution for the given model. If missing, a default value is computed based on FBA. If given, arguments solver and method are used to calculate the dafault, but solverParm is not.

delta

A single numeric value giving the relative range of tolerance, see Details below.
Default: 0.03.

epsilon

A single numeric value giving the absolute range of tolerance, see Details below.
Default: 0.001.

LPvariant

Boolean. If TRUE, the problem object is formulated as linear program. See Details below.
Default: FALSE.

LPvariant

Boolean. If TRUE, the problem object is formulated as linear program. See Details below.
Default: FALSE.

absMAX

A single numerical value used as a maximum value for upper variable and contraint bounds.
Default: SYBIL_SETTINGS("MAXIMUM").

cnames

A character vector giving the variable names. If set to NULL, the reaction id's of model are used.
Default: NULL.

rnames

A character vector giving the constraint names. If set to NULL, the metabolite id's of model are used.
Default: NULL.

pname

A single character string containing a name for the problem object.
Default: NULL.

scaling

Scaling options used to scale the constraint matrix. If set to NULL, no scaling will be performed (see scaleProb).
Default: NULL.

writeProbToFileName

A single character string containing a file name to which the problem object will be written in LP file format.
Default: NULL.

...

Further arguments passed to the initialize method of sysBiolAlg. They are solver, method and solverParm.

The problem object is built to be capable to perform the ROOM algorithm with a given model, which is basically the solution of a mixed integer programming problem

% \begin{array}{rll}% \min & \sum\limits_{i=1}^n y_i \\[2em] \mathrm{s.\,t.} & \mbox{$Sv$} = 0 \\[1ex] & \alpha_i \leq v_i \leq \beta_i & \quad \forall i \in \{1, \ldots, n\} \\[1ex] & v_i - y(\beta_i - w_i^u) \leq w_i^u \\[1ex] & v_i - y(\alpha_i - w_i^l) \geq w_i^l \\[1ex] & y_i \in \{0, 1\} \\[1ex] & w_i^u = w_i + \delta |w_i| + \epsilon \\[1ex] & w_i^l = w_i - \delta |w_i| - \epsilon \\[1ex] \end{array}%

with \bold{S} being the stoichiometric matrix, \alpha_i and \beta_i being the lower and upper bounds for flux (variable) i. The total number of fluxes of the optimization problem is denoted by n. Here, w is the optimal wild type flux distribution. This can be set via the argument wtflux. If wtflux is NULL (the default), the wild type flux distribution will be calculated by a standard FBA. All variables y_i are binary, with y_i = 1 for a significant flux change in v_i and y_i = 0 otherwise. Thresholds determining the significance of a flux change are given in w^u and w^l, with \delta and \epsilon specifying absolute and relative ranges in tolerance [Shlomi et al. 2005].

The Boolean argument LPvariant relax the binary contraints to 0 \leq y_i \leq 1 so that the problem becomes a linear program. The optimization can be executed by using optimizeProb.

Objects from the Class

Objects can be created by calls of the form

sysBiolAlg(model, algorithm = "room", ...).

Arguments to ... which are passed to method initialize of class sysBiolAlg_room are described in the Details section.

Slots

wu:

Object of class "numeric" containing the upper threshold for a significant flux change, see Details below.

wl:

Object of class "numeric" containing the lower threshold for a significant flux change, see Details below.

fnc:

Object of class "integer" containing the number of reactions in the entire metabolic network (argument model to the constructor function sysBiolAlg).

fnr:

Object of class "integer" containing the number of metabolites in the entire metabolic network (argument model to the constructor function sysBiolAlg).

problem:

Object of class "optObj" containing the problem object.

algorithm:

Object of class "character" containing the name of the algorithm.

nr:

Object of class "integer" containing the number of rows of the problem object.

nc:

Object of class "integer" containing the number of columns of the problem object

fldind:

Object of class "integer" pointers to columns (variables) representing a flux (reaction) in the original network. The variable fldind[i] in the problem object represents reaction i in the original network.

alg_par:

Object of class "list" containing a named list containing algorithm specific parameters.

Extends

Class "sysBiolAlg", directly.

Methods

optimizeProb

signature(object = "sysBiolAlg_room"): runs optimization on the given problem object (see optimizeProb for details).

Note

If using glpkAPI as MIP solver, consider to set parameter PRESOLVE to GLP_ON.

Author(s)

Gabriel Gelius-Dietrich <[e-mail]>

Maintainer: Mayo Roettger <[e-mail]>

References

Shlomi, T., Berkman, O. and Ruppin, E. (2005) Regulatory on/off minimization of metabolic flux changes after genetic pertubations. PNAS 102, 7695–7700.

See Also

Constructor function sysBiolAlg and superclass sysBiolAlg.

Examples

  showClass("sysBiolAlg_room")

SysBioChalmers/sybil documentation built on Aug. 24, 2023, 1:08 p.m.