Potential Error in fluxVariability.m

[arichelle]

When we perform a flux variability analysis only on a set of reactions (and not on the complete model), there might a problem with lines 237, 238 maxSolved = false(size(model.lb)); minSolved = false(size(model.lb)); Indeed at this point of the code, there is a solution to the problem and we try to compute the boundaries of the specified fluxes. But these 2 lines do not take into account that the analysis is only performed on a set of reactions. Proposition of change is maxSolved = false(size(Order(Presence)))); minSolved = false(size(Order(Presence)))); Doing so, there is no more problem of dimension in the computed solution.

This should be double checked ;)

[tpfau]

Thats entirely possible, even though we use reaction lists during the tests. Can you write a test where the results using the current version are wrong so that we can debug it properly?

[shjchan]

I believe this problem has been solved now.