Questions about running ANCOM-BC at the ASV level

[rylee650]

After referring to a few forum posts on ANCOM-BC, I decided to run ANCOM-BC at the ASV level in addition to using my table collapsed at the L7 level for my fungal ITS data. After assigning taxonomy to the individual feature IDs in R I noticed some of my treatment groups have the exact same taxa enriched with different log fold change values.

The issue I am having is how to properly visualize this data for a presentation. I had made a heat map in R to show the magnitude change of each taxa, but I am not sure how to go about this for the ASV level. Should I just keep all ASVs in individual rows, or if there some way to combine the taxa to get one LFC value?

I appreciate any insight as I am still relatively new at this!

[timanix (Timur Yergaliyev)]

Hello!

Usually, I check if my results with collapsed data agree with the test at the ASV level. If yes, I just use output from collapsed test. If ASVs behave differently from taxa (ASVs from the same taxa have the opposite signs of LFC values, or just too different), then I use first 5-6 characters from ASV Id and taxonomy (Lactobacillus_abc123, Lactobacillus_def456).