I recently made a heatmap and had a few quick questions about it.
Is there a setting where I can put the names of the phylums in alphabetical order? Also, I grouped my samples together by age prior to running the command and had the same question about them except if I can put them in numerical order? The reason I ask is that I’m trying to compare multiple heatmaps to each other but it’s difficult when the taxa aren’t in the same order. I’m also doing a longitudinal comparison within a heatmap but my groups are out of order so it’s difficult to view the changes.
Also, I tried to print my heatmap out but I don’t think it’s size is being recognized (I’m not sure if this is an issue with my computer, printer, or the Qiime2 SVG) so when I print it, it’s only printing about a third of the size on one paper and the other two thirds never print out. Is there a way to fix this?
Short answer you can disable clustering on one axis, or the other, but not both at the same time. We have an open issue for that. Clustering is what causes the IDs to render in a non-sorted order (makes sense, right?)
Sounds like something downstream of QIIME 2 - as a quick double-check I just generated and printed a heatmap, all worked as expected (I did need to make sure my print settings were set to “scale” in order to fit the whole thing on one page…).
So then I have to use seaborn to do this? Qiime2 does not have any code for clustering? For using seaborn, would my input be the same as the one I put into the Qiime2 heatmap code? Thanks!
As for scaling, I’ve tried that before and it comes out so small that it is no longer legible. I am creating my heatmap using level 6 taxonomy so it’s too large to try and fit on one page.
No, the exact opposite - QIIME 2 only supports clustering right now, there is no option to disable clustering on both axes - which is what you asked for above.
You could - I was linking to that because that is the code that q2-feature-table is referencing within the heatmap method.
Maybe printing isn't the right solution for understanding these data? Perhaps you could consider grouping your feature table along the feature axis to consolidate some features.