Features of the flux optimization in FASIMU

The following is just a selection, for full reference, please see the manual.
The main call for a single flux optimization is "compute-FBA" and it has the following options:

-F <fluxmin-type>

type of the fluxmin computation

0: backward and forward flux weighted equally
1: backward flux weighted with equilibrium constant
2: backward flux weighted with equilibrium constant, quadratically adjusted
s: backward flux weighted with equilibrium constant, adjusted Scott Holzhütter method (Holzhütter & Holzhütter, 2004)
e: flux weigthed upon enzyme costs
E: flux weigthed upon enzyme costs, backward flux additionally weighted with equilibrium constant

-m <rea-names> ...

Minimization of these reactions (several rea names may follow)

-b <rea-names> ...

Maximization of these reactions (several rea names may follow)

-f [<type> [<weight>]]

use fitness maximization instead of fixing targetfluxes, default type 3, default weight 1e5

0: unadjusted type
1: linearly adjusted type
2: fully adjusted type (quadratic)
3: Computationally very easy normalized linear type, but the value of the targetflux will be restricted to the same sign as the target value
4: linear type. Not normalized.
5: linear type. Normalized.
6: maximum norm, Not normalized.
7: maximum norm, Normalized.
Normalization:: each component is divided by the (squared for 0-2) absolute value of targetflux. Ensures that each targetflux has the same relative impact. If the targetflux is zero, its impact is considered to be 1.

Options for expression:

--xs [<lower threshold> [<upper threshold> [<significant flux> [<weight>]]]]

Shlomi-type expression match

<lower threshold>: lower threshold on the expression level, default 0.5
<upper threshold>: upper threshold on the expression level, default = lower threshold
<significant flux>: threshold on significant flux, default 1
<weight>: weight relative to other cost factors, default 1e3

-xg [<threshold> [<weight>]]

GIMME expression match (PMID:18483554)

<threshold>: threshold on the expression level, default 0.5
<weight>: weight relative to other cost factors, default 1e3

Options for thermodynamic realizability (TR) constraint:

-T

use TR criterion, default type

-T [A-Ea-e]r?

use TR with type identifier

-k <number>

potential tolerance maximum

-K <number>

penalty multiplicator for potential tolerance

-s [<weight>]

use soft bounds, set penalty weight (default 1)

-w [<weight>]

use set points, to be found in concentration ranges, default weight 1e-5

concentration range if no file concentration-ranges is available

Program control options:

-p

prepare only 'problem.lp', do not compute

-t <seconds>

timeout in seconds, finish (single flux computation), default 300, 0 means unlimited time

-CPLEXnodes <nodes>

abolish computation number of nodes, default 5000, 0 means unlimited number of modes

-CPLEXemph <1..4>

MIP emphasis setting to be tried out in critcal size models

-CPLEXline "<cplex-parameter-line>"

Give this line to CPLEX

-c

check the results with another optimizer run

-d

print documentation files

More options are controlled by the following files:

fluxfix: fixes the flux of one reaction to a linear combination of other fluxes
expressions: expression data
enzymes: cost of specific enzymes, to be used in conjunction with -F e or -F E
fluxmin-excluded: list of reactions having no weight
TR-excluded: list of reactions excluded from TR criterion
concentration-class-ranges: Range of concentration values for a concentration class, given as: lower hard bound, lower soft bound, set point, upper soft bound, upper hard bound
concentration-ranges-classes: Assigns a metabolite to a concentration range