ANCOM output - not sure how to interpret lack of results


I have tried to do differential abundance testing with ANCOM for two different clinical groups (ex: patients who gained vs. did not gain weight). However, I get a volcano plot that doesn't show up. So, I'm not sure if I'm doing anything wrong or how should I interpret this?

Attached is the output file as well as the comp-timepoint2-table.qza, and a part of the acutal sample-metadata.tsv file (the actual metadata file that I had ran is much larger).

Thanks so much for helping out!

sample-metadata-forancomquestion_timepoints.tsv (1.8 KB)

l4-ancom-new_wt_chng_category_BisLoss.qzv (34.7 KB)
comp-timepoint2-table.qza (40.9 KB)

Hi @rnasrah - I’m looking at your results. The volcano plot is not showing up, mainly because there are no differentially abundant microbes identified.

Question: Are you able to see differences in the PCoA plots, or other beta diversity measures (i.e. PERMANOVA?). If you aren’t able to see differences in there, you are probably not going to be picking up differentially abundance species.

It may also be possible that ANCOM is just being over conservative, and is throwing out differentially abundant microbes. It is possible to tune ANCOM to be more sensitive - that can either be done using the scikit bio function or in the original ANCOM R package.

Alternatively you could try out gneiss to see if there are aggregates of microbes that could be explanatory, rather than individual microbes.

1 Like

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.