Problem reproducing secreted metabolites in solution.

[lkroeh]

Hi,

I'm using the human RECON2 model, where I load in the model, change the objective function to biomass, change the media, and then do solution = model.optimize(); followed by print(model.summary(solution=solution)). I see lactate is secreted in the summary.
When I do solution.fluxes.EX_lac_L_b (the reaction for lactate production) I can confirm there is a flux.
However, when i do model.metabolites.get_by_id('lac_L_b').summary(), there is no information reported!

Additionally, when I re-run print(model.summary(solution=solution)), I not longer see lactate as a secreted product, although the biomass solution value does not change.

So 2 questions:

1. Why when I see lactate secretion reported in the summary, can I can not find information on what reaction is producing it?
2. Why are the secreted products changing when I re-run the code chunk which defines the media (but without changing any values from the run before) and print the solution?

Thanks for your help!!
Lina

[Midnighter]

Hi Lina,

Without having looked at it, my guess is that this is due to there being many different flux distributions that enable the same optimal objective value (non-unique solution). You can confirm this by also using FVA in the summaries. You should see that there many potential candidates for being secreted, not only lactate.

You can get a more reproducible solution by using parsimonious FBA, although that still does not necessarily give you a unique solution. Do you just want to understand the model better or do you have a particular biological question that you would like to answer/model?

[lkroeh]

Hi, Thank you Midnighter!! You're very right, I will use FVA/pFBA. I just want to understand the model better, but am also now intrigued by the lactate production that I sometimes see, so also want to investigate why that may be happening.
Best,
Lina