feat/anaerobicgrowth&Nitrogen-limited growth

[feiranl]

Main improvements in this PR:

• To update the function anaerobicModel.m to fix the issue mentioned here fix: zero-growth in anaerobic conditions #198. heme was removed as in earlier function. This time, we update the function by removing heme, NAD, NADP, NADPH and NADH when we simulate anaerobic condition.
• To update the Nitrogen-limited condition growth by rescale the protein content
• To update a growth test for model validation, which is also the figure S4c for the yeast8 paper.

I hereby confirm that I have:

• Tested my code with all requirements for running the model
• Selected devel as a target branch (top left drop-down menu)
• If needed, asked first in the Gitter chat room about this PR

[BenjaSanchez]

@feiranl after the refactoring changes I introduced I don't have any more issues, we can merge :)

• There's a new function scaleBioMass which can be used for rescaling any biomass component, but without the additional changes from the old script. It also has an option for balancing out the total mass so that it adds to 1 g/gDW. The old scaleBioMass has been renamed to fixBiomassComp to be more explicit about its purpose (does not break BC as the old scaleBioMass was not a function to begin with).

• sumBioMass does not rely on the data input anymore, as it builds it from within.

• rescalePseudoReaction has been isolated so that it can be called by its parent but also by the new scaleBioMass.

• I noticed that predictions changed a bit, and this is I believe due to what we discussed before: For example, in the case of N-limited anaerobic conditions, with the old scaleBioMass it didn't matter the changes you were introducing in the nitrogen section, as they were being over-written by the original changes when calling anaerobicModel (which called scaleBioMass again). See below the old code:

  if strcmp(mode2,'N')
      content = {'carbohydrate','protein','lipid backbone','RNA'};
      fraction = [0.587 0.289 0.048 0.077];
      model = scaleBioMass(content,fraction,model,false,false);
  end
  if mode1 == 2
      model = anaerobicModel(model);
  end

  This is now not an issue anymore, however predictions do look slightly worse:

  

Let me know if the code looks good on your end (especially if the biomass composition for anaerobic/N-limited/etc conditions is being properly built), or if you have any additional comments.

[hongzhonglu]

Nice work!

[feiranl]

It looks good from my side. I run the code on MAC, it generate the figure perfectly.