MetaCyc pathways from unclassified species

[Nick]

Do genes have to come from the same species in pathway identification? For example, Pathway X requires Gene 1 and Gene 2 to come from Species Z to count as a valid pathway hit?

What happens if Gene 1 is from an unclassified species and Gene 2 is from another unclassified species. Would Pathway X still show up in the results? 

[Eric Franzosa]

Hi Nick,

Community-level pathways (i.e. the ones in the output that aren't followed by "|species" or "|unclassifed") are quantified from community-total abundances. Hence, they could represent a mixture between components from different species, known or unknown. Pathways in the "unclassified" stratum are similarly computed from "unclassified" reaction abundance, which could be coming from one or many species.

Thanks,

Eric

On Tue, Oct 15, 2019 at 2:26 AM Nick <nicochun...@gmail.com> wrote:

Do genes have to come from the same species in pathway identification? For example, Pathway X requires Gene 1 and Gene 2 to come from Species Z to count as a valid pathway hit?

What happens if Gene 1 is from an unclassified species and Gene 2 is from another unclassified species. Would Pathway X still show up in the results? 

--
You received this message because you are subscribed to the Google Groups "HUMAnN Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to humann-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/humann-users/f4985fe7-71b7-49db-aa5e-2ca48a253ece%40googlegroups.com.

[Nick]

Hi Eric,

   Thanks, just wanted to make sure I wasn't misunderstanding. So if we want to be more conservative and only look at which pathways exists within a single species, we should only consider PathwayX|speciesY. 

For community-level and unclassified pathways, does it make from an ecological perspective to construct pathways this way? In other words, would two separate genes expressed by two separate species ever form a pathway in practice? I know many species utilize the metabolites of other species to form certain compounds, but I'm sure there are many reactions that have to occur with a single native cell.

Thanks,

Nick

On Wednesday, 16 October 2019 00:10:42 UTC+8, Eric Franzosa wrote:

Hi Nick,

Community-level pathways (i.e. the ones in the output that aren't followed by "|species" or "|unclassifed") are quantified from community-total abundances. Hence, they could represent a mixture between components from different species, known or unknown. Pathways in the "unclassified" stratum are similarly computed from "unclassified" reaction abundance, which could be coming from one or many species.

Thanks,

Eric

On Tue, Oct 15, 2019 at 2:26 AM Nick <nicochun...@gmail.com> wrote:

Do genes have to come from the same species in pathway identification? For example, Pathway X requires Gene 1 and Gene 2 to come from Species Z to count as a valid pathway hit?

What happens if Gene 1 is from an unclassified species and Gene 2 is from another unclassified species. Would Pathway X still show up in the results? 

--
You received this message because you are subscribed to the Google Groups "HUMAnN Users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to humann...@googlegroups.com.

[Eric Franzosa]

   Thanks, just wanted to make sure I wasn't misunderstanding. So if we want to be more conservative and only look at which pathways exists within a single species, we should only consider PathwayX|speciesY. 

That's correct. Or you might focus on pathways whose community totals are similar to the sum of individual species' contributions.

 

For community-level and unclassified pathways, does it make from an ecological perspective to construct pathways this way? In other words, would two separate genes expressed by two separate species ever form a pathway in practice? I know many species utilize the metabolites of other species to form certain compounds, but I'm sure there are many reactions that have to occur with a single native cell.

It's grounded in earlier approaches to functional profiling that didn't attempt to discern species boundaries and instead focused on the community as a "bag of genes" that satisfied a set of pathways. I think this is still a useful approach for defining an "upper bound" on the space of pathways that a community can satisfy. Conversely, treating community pathways as ONLY the set that can be satisfied by individual species would probably lead to an underestimate as it neglects the potential for reactant export/import.

Thanks,

Eric