Why has GNEISS been deprecated from Qiime2?

[ecg (Ele)]

I am trying to re-run an analysis I had from a couple of years ago, and I see https://docs.qiime2.org/2020.11/tutorials/gneiss/ that gneiss has been deprecated in the newer version of QIIME, why is this? and if I’d like to run gneiss on R, how can I import my gneiss_hierarchy.qza and gneiss_heatmap.qzv into R?

Thanks,

[mortonjt (Jamie Morton)]

Hi @ecg, thank you for your interest in gneiss.

Gneiss has not been deprecated – only some of the visualizers have been deprecated in favor of Empress. By the next release, there will be compatibility between Songbird (possibly Aldex2) and Gneiss, where the ILR transform can be applied directly to differential abundance analysis results.

Of course, there will be more features coming up in the future (hence the FeatureTensor discussion)
See the following pull requests.

adding ILR figure

Ilr ordination

Regarding your questions

1. gneiss_hierarchy.qza can export a newick file, which can be read into R. See https://docs.qiime2.org/2019.10/tutorials/exporting/#exporting-a-phylogenetic-tree
2. gneiss_heatmap.qzv, this probably can’t be ported into R, but there are a number of tools that support phylogenetic heatmaps in R. See for instance https://rdrr.io/cran/phytools/man/phylo.heatmap.html

[TurboQiimer]

Hi @mortonjt
Is GNEISS alternative software for ANCOM? Am I right? And among the different GNEISS methods, which one is widely used and which one do you suggest?
Thanks
Qiimer

[Mechah]

@mortonjt

By the next release, there will be compatibility between Songbird (possibly Aldex2) and Gneiss, where the ILR transform can be applied directly to differential abundance analysis results.

Awesome, thanks!

[ecg (Ele)]

Looking forward to the next version! Thanks @mortonjt