Humann_regroup table - uniref90_rxn vs uniref90_level4ec


I have a small confusion about running humann_regroup table.
I successfully executed humann (version3), normalised the results of my run using humann_renorm_table and concatenated the output using human_join_tables.

Looking at the help file of humann_regroup table I see that the only options that I have are uniref90_rxn and uniref50_rxn.

Can you tell me it there is a difference between uniref90_rxn and uniref90_level4ec?
If yes can you explain what it is and how can I obtain uniref90_level4ec?

Thank you very much in advance for your time.

PS: can you also explain what humann_reduce_table does exactly as I can not understand if i need to use it or not? I could open another topic for that if you like.

  • RXNs (MetaCyc reactions) and ECs are very similar; most proteins are associated to reactions via their EC annotation. However, there are additional proteins that are associated directly to reactions in MetaCyc, leading to differences between the two mappings. We use reactions for computing pathway abundance but I tend to use ECs for other applications since they are more generic.

  • It sounds like you need to download the HUMAnN utility mapping database to get these other mapping files. Please see the manual for instructions.

  • Reduce table is used for filtering rows of a table in certain applications (as outlined in the manual). It is not one of the more commonly used helper scripts.

Thank you very much for the quick reply. I am downloading the utility mapping db now so I ll let you know if I have any problems.

Can I ask what is the filtering criterion in the reduce_table script? I could not find a detailed description in the manual.

Thx a lot