Some questions about the targets

[pablo-roman-escriva]

Hi,

We are trying to run an analysis in the context of microbiota and symbiosis between species. For that, we have a set of annotated genomes, a medium(gut seeds from the publication), and a set of targets that we would like to reach.

The first question concerns the compartment codification within the target's ID. To study the organisms' metabolic cooperation, we assumed that the list of reachable targets must be exported extracellularly so that another organism could incorporate them. To achieve this, the compartment codification of the targets should be M_id_e instead of M_id_c or does the program take this into account independently of if the target is specified as intracellular (c) or extracellular(e)?

The second question is about the name of the compounds themselves. For example, If we would like to use lactate as a target, could we use LACTATE (compound class) or should we use both D-LACTATE and L-LACTATE instead?

Thanks,

Pablo

[cfrioux]

Hi Pablo,

Transport reactions may not be well annotated or precise enough to make the assumption that reachable targets will be exported. Additionally, some transporters are non-specific, passive transport can happen, and cell lysis as well. That is why we decided not to explicitly account for transports but rather overestimate the potential for exchanges, even without annotated transport in the networks.

If you set up extracellular metabolites as targets and no transport reaction exists in any of the metabolic networks, targets will not be reached/producible. I would stick with the cytosolic version first, ensuring that the metabolite is indeed producible, and eventually check the transport with dedicated tools.

Regarding the second question, that is Metacyc-dependent: no reaction connect the three lactate compound so picking the wrong one may have a negative effect on modeling. You should use the ID that is present in your metabolic networks. Safest way to proceed would be to add the three versions of lactate. Nothing bad will happen if you add an orphan compound: outputs of M2M (json files) will tell you which seeds are not used because absent from metabolic networks.

Let me know if something remains unclear.

Clémence

[pablo-roman-escriva]

Thank you for your reply. It has been really helpful.