Mapping MetaCyc Pathways

Microbial community profiling HUMAnN
1

Hello,

I’m interested in particular MetaCyc pathways, but they don’t show up in the pathabundance or pathcoverage results. My pathways of interest are missing. Is there a way I can look for specific pathways in my shotgun sequences?

Thank you

2

I tried turning MinPath off and re-running my samples, but the pathways are still missing.

It’s odd because the genes from the pathway show up in the genefamilies file, EC numbers, but the corresponding pathway does not.

Any help would be appreciated.

3

Hello, I have the same problem. I am trying to figure out better what happens with the pathway calculations…

4

To quantify a pathway we have to be able to find MOST of its reactions/ECs in the sample with non-zero abundances. So it’s possible you would see some ECs from the pathway but not see the pathway itself if a lot of the important reactions were missing. If you feel like most of the reactions/ECs are non-zero and you’re STILL not seeing an abundance for the pathway, then please let us know (as this might be some other issue).

5

Hello franzosa, and thanks for your help,

Let me explain a little better my problem.I am interested in the pathway that creates GABA. That would be meta-cyc PWY0-1305, which is a product of only one gene called glutamate decarboxylase. Now I find this gene with other functional annotation pipelines (e.g. KOfam scan) and I find this gene also through prokka in MAGs at high abundance in my sample of interest. Nevertheless, the pathway with humann does not show up.

Also how can I look for a specific gene in the humann gene annotation? should I get all the uniref90 IDs associated to that gene from the uniprot website and just grep my genefamilies result table?

Thanks for the help and the cool tool you developed,
Gabri

6

Thanks for the clarification. HUMAnN only quantifies pathways that contain at least 4 reactions. If you’d like to quantify a single reaction, you can regroup (sum) your UniRef90 abundances to ECs or MetaCyc reactions with the humann_regroup_table script and look up the reaction of interest (looks like 4.1.1.15 in this case?).

2 Likes