ROU optimization

[Laminsecin]

Hi,

I wanted to ask what the numbers in front of reactions in the modification part actually mean?

[vmspereira]

Hi @Laminsecin

Over (under)-expressions are modeled by changing the reactions' lower (upper) bound to a factor of a reference distribution of flux values. The numbers in front of each reaction are the applied factors (folds).

Let us suppose, for example, that the reference flux rate value for the R_ACGH reaction is 0.1 and that the reaction bounds (irreversible reaction) are $$0 \leq R \textunderscore ACGH \leq 1000.$$ The 32, meaning an overexpression of 32 folds, will modify the reaction bounds to $$32\times 0.1 \leq R\textunderscore ACGH \leq 1000,$$ forcing a higher flux rate for the R_ACGH reaction. Down regulations, with factors in the ]0,1[ range, will modify the reaction upper bound, causing a reaction flux rate lower than the reference one.

The user can provide the reference flux distribution as a dictionary of reaction_id, flux value mapping. If none is given, the flux distribution obtained when running a parsimonious FBA for the model objective will be considered. The fold values are configurable and may be any discrete list of nonnegative values.

# folds
levels = [0.1 ,0.2 ,0.3 ,1.1 ,1.2, 1.3]

reference = {'R_ACGH':0.1, ... }
problem = ROUProblem(model, levels=levels, reference=reference)

You may find additional information on the MEWpy documentation.

[Laminsecin]

@vmspereira Thanks a lot for your response.