Utility_mapping full

I got the metabolic pathways through Humann2, but I am interested in making a map of all important pathways. Please, can someone suggest to me how I can visualize themes using “Pathway Collages"? Pathway Collages

And also, how can I map my data against KEGG or any other pathway database? I already downloaded the utility_mapping files with the following command: humann_databases --download utility_mapping full $humann_out so if you have any simple code for that, I appreciate it

the third question is: how can I combine the taxonomic profiling and the function information in one table? (I know how can do it via Galaxy but I need a simple command.)

thanks a bunch
Mohamed S. AboHoussien

Where MetaCyc is now behind a subscription I can’t offer a lot of guidance on how to generate pathway collages. We are looking into alternative pathway systems for future releases.

HUMAnN can natively regroup UniRef90s/50s to KEGG Orthogroups (KOs) because this information is provided freely by UniProt. We do not bundle the KEGG pathway definitions because doing so requires an academic license. If you have access to KEGG, the pathway/modules are defined there using a language HUMAnN understands (indeed, HUMAnN’s format was based on the format originally implemented by KEGG).

I’m not sure what you mean about combining the taxonomic and functional information? If you mean including taxa rows from MetaPhlAn and function rows from HUMAnN, I would not recommend that (too easy to get confused).

1 Like

Thank you Erick for your kind response.
so I can’t for example make signature pathways related to each other or visualize it anyway?

KEGG pathways need academic licenses what about GO for example or other databases?

Using GO terms does not require a license (thankfully!). We bundle a mapping from UniRef90 to GO in the utility mapping database. We provide all of the mappings available in UniProt, so you would need to select the terms you are most interested in, and potentially bubble up mappings through the GO hierarchy. For example, if UniProt associates gene X with term Y, it can be inferred that gene X is also associated with Y’s parent term(s) also.

Then how can I map my data against the GO database? if you have any simple code for that, I appreciate it

Thank you Erick it’s already done