Leo Lahti

Is this critical or can we close?

Then let's just close it.

This could go, for instance, in ch. 15 (visualization) / heatmaps.

I agree that sechm is good for neatmaps and ComplexHeatmap for hclust. In fact ch 15.3 could be simplified in this regard. - explicitly mention that sechm does neatmap (and...

This topic is now under construction in miaViz PR https://github.com/microbiome/miaViz/pull/128 - in fact I forgot we had this issue.. Let us finalize this OMA issue after the miaViz PR is...

1) If neatsorting is not available in SEtools then we can forget about that 2) Could you show the code that does this?

For final examples in the R documentation it will better to have just once the parts in data preparation that are common for both methods (agglom, transf). Code 1: I...

Usually we would recommend the following transformations for this kind of heatmap: - clr (or log10p) for samples - z (or now, "scale") for features The reason: clr will logarithmize...

Yep I realized afterwards. I think those transformation updates might help

CLR can be done for both counts or relabundances. If you use CLR + Euclidean distance, then PCA would be the common option. Otherwise, relabundance + Bray-Curtis -> MDS /...