I did a recent study using QIIME2019.7. The target region was V4, with samples of human faecal microbiome, and I used the gg-13-8-99-nb-classifier.qza from the QIIME2 website for taxonomy. There are 12 samples in this study, with 5 in one treatment and 7 in the other. I have a few queries please:
- I do understand that there can be multiple entries for the same taxon. e.g. There are 40 entries for k__Bacteria; p__Firmicutes; c__Clostridia; o__Clostridiales; f__Lachnospiraceae in this table.MicroAnal_RLSfilt.csv (125.2 KB) However, I want to do draw total and rel abundance graphs for each treatment at different taxa levels. What is the easiest way to do this please? Would it be through the taxa collapse function? If so, what are the steps?
Then using a filtered table (removing frequencies under the median, and then below 0.1 %), I did ANCOM and Gneiss analysis.
- I did an ANCOM (to see what comes out in the wash as one of previous posts suggests); with no significant features being detected. filt-ancom-Group.qzv (486.3 KB)
- I also tried gneiss (add pseudocount->correlation clustering->ilr-hierarichal->ols-regression.). Based on the FDR corrected p value, I really did not see much happening, but I chose balances for visualization y25 (p=0.051) and included y0 (p =0.99) for comparison. I am not sure what to interpret and would appreciate some pointers please. Also attached is the regression summary- Reading through the post and the links within, I am not sure there is much happening, but appreciate if you can suggest what I might interpret.filt-regression_summary_Group.qzv (677.0 KB) filt-y0-L-1-Group-summary.qzv (151.4 KB) [filt-y0-L-6-Group-summary.qzv|attachment]filt-heatmap_group.qzv (184.0 KB) (upload://fTHINKkYJ36cO4Am0sYBTLfytY9.qzv) (151.2 KB) filt-y25-L-1-Group-summary.qzv (128.0 KB) filt-y25-L-6-Group-summary.qzv (129.4 KB)
- What else can I do, just to see trends more than find definitive statistical differences?