Question Regarding predictMicrobeContributions()

[Aarush Garg]

Hello,

I’m seeking clarification regarding thepredictMicrobeContributions() method (as implemented in the COBRA Toolbox/mgPipe, for example in Heinken et al., Microbiome 2019).

Background:

• After generating diet-constrained microbiome models using mgPipe, each individual model’s objective is set to the community biomass (e.g., EX_microbeBiomass[fe]).

• When predicting microbe-specific contributions to metabolite exchange using FVA (predictMicrobeContributions.m), should the model.objective be left as the biomass reaction (i.e., as set in the generated model), or should it be changed before running FVA on exchange reactions?

• My understanding—supported by both the code and the aforementioned paper—is that the intention is to maintain the biomass objective. Thus, FVA computes min/max fluxes through each exchange reaction while requiring (near-)optimal growth via the fraction of optimum constraint. Is this correct?

My PI is unsure whether FVA should constrain on biomass or if the objective should be set to something else for this function when trying to find strain-level contributions and wants confirmation this matches published workflows.

Specifically, we were looking at Ines Thiele IBD Paper, and based off the Materials and methods/Interrogation of models for bile acid synthesis capabilities, wonder if the optimum does indeed remain as the EX_microbeBiomass[fe] reaction.

Any confirmation or pointer to documentation/published discussion would be greatly appreciated!

[Almut Heinken]

Hello,

by default and in the mentioned publications, the predictions of microbe-metabolite contributions is indeed performed while assuming community biomass as the objective function with flux in the set range (0.4-1 mmol/person/day). However, if you think another objective function and/or set range of allowed community biomass production is more appropriate for your application, you are free to change this. This modification would have to be done before starting the predictMicrobeContributions functino, which loads the models from the folder in their current state.

I hope this helps,

best regards,

Almut

--

---
You received this message because you are subscribed to the Google Groups "COBRA Toolbox" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cobra-toolbo...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/cobra-toolbox/a1ef5486-f65e-4004-bb21-36fb206122b4n%40googlegroups.com.