These are chat archives for opencobra/cobrapy

18th
Jul 2017
Bhushan Dhamale
@BhushanDhamale
Jul 18 2017 06:36
Guys, I've got a couple of questions.
  1. How are the reactions contributing to the biomass reaction determined?
  2. How is c (the vector of weights) determined for every reaction contributing to the biomass reaction?
Henning Redestig
@hredestig
Jul 18 2017 09:22
@BhushanDhamale on 1, do you mean scientifically or practically in cobrapy? for the latter, checkout model.reactions.<biomass_reaction>.reactants which gives you a list of metabolites and for each of those you can fetch its corresponding reactions
Bhushan Dhamale
@BhushanDhamale
Jul 18 2017 10:31
@hredestig I meant practically in COBRApy, but it would be nice to know the scientific basis for it as well.
Bhushan Dhamale
@BhushanDhamale
Jul 18 2017 10:42
model.reactions.<biomass_reaction>.reactants is just giving me a list of reactants taking part in the biomass reaction.
I want it to display the c value for every reactant in the reaction, and I need to know how it is calculated for every reactant in the equation.
Henning Redestig
@hredestig
Jul 18 2017 11:10
that's part of model building, best check the original paper for your model of interest I guess
Bhushan Dhamale
@BhushanDhamale
Jul 18 2017 11:32
@hredestig Will do! :+1:
Thanks for the help Henning. :smile:
Bhushan Dhamale
@BhushanDhamale
Jul 18 2017 12:42
Script
After deleting the PFK reaction, I am attempting to calculate the new absolute predicted flux by FBA and pFBA in the Salmonella test model. However, I am getting the same values as before the KO.
How do I re-calculate the absolute flux values?
The output of the script appears so,
STM PFK before KO:  0.0 < pfk < 1000.0
# Absolute predicted flux before PFK KO
1. By FBA   :  448.6754107129603
2. By pFBA  :  343.0211601643632
Diff. betn FBA and pFBA :  105.65425054859708

STM PFK after KO:  0 < pfk < 0
# Absolute predicted flux after PFK KO
1. By FBA   :  448.6754107129603
2. By pFBA  :  343.0211601643632
Diff. betn FBA and pFBA :  105.65425054859708
Henning Redestig
@hredestig
Jul 18 2017 12:55
you need to recompute your stm_fba and stm_pfba after the knock-out. Also, in that script you don't even create those objects to begin with so something is missing there..
Bhushan Dhamale
@BhushanDhamale
Jul 18 2017 16:28
@hredestig My bad! Actually, it's a poorly patched code from my jupyter notebook. I had defined them in a separate block and forgot to include them in the gist. :sweat_smile:
Bhushan Dhamale
@BhushanDhamale
Jul 18 2017 16:47

There! I've updated the gist.
However, it gives me,

STM PFK before KO:  0.0 < pfk < 1000.0
# Absolute predicted flux before PFK KO
1. By FBA   :  448.6754107129603
2. By pFBA  :  343.0211601643632
Diff. betn FBA and pFBA :  105.65425054859708

STM PFK after KO:  0 < pfk < 0
# Absolute predicted flux after PFK KO
1. By FBA   :  525.8982922803317
2. By pFBA  :  343.0211601644425
Diff. betn FBA and pFBA :  182.8771321158892

Intuitively, one would expect the flux to decrease after knocking out a reaction; here, I'm getting an increase in the absolute predicted flux after knocking out the PFK reaction. Is it possible that other reaction pathways might be compensating for the loss of the PFK reaction, leading to an overall increase in absolute predicted flux by FBA?

Christian Diener
@cdiener
Jul 18 2017 17:34
Yes very likely that is happening. Since cutting PFK also means cutting glycolysis might be that you have to import more things from the environment to sustain the optimal growth rate.
Also for pFBA you should try to constrain the imports correctly otherwise pFBA often uses a solution that imports everything from the outside
shutting down most of the internal metabolism