Aaron Lun

Your output looks fine to me. Consider providing a minimum reproducible example.

You'll find that `sceM$label` is `NULL`.

The problem seems to be with `geom_tile`. Here is a minimum working example: ```r library(shiny) library(ggplot2) ui

A separate assay would be trivial, even for HDF5. Set it up as a DelayedArray and you can get a free log2 transformation that is present without actually performing the...

Onto the actual questions: > do point selections behave well in the transformed space? I don't even know what this means. A selection is a selection in whatever space it...

The rectangular selections would be fine for monotonic transformations. It's true that the lasso might do weird things with respect to the convex hull, because the user will think that...

`.generateDotPlot` is contractually obliged to return a dot plot in the `dot.plot` variable. One could add a command to create a unique variable, but that means that subclasses that `callNextMethod()`...

> Now one question, is there an easy way to `coord_flip` my feature assay plot, ie have tissue on the y-axis and the gene expression on the x-axis ( =...

cannot reproduce.

In the meantime, you could consider chucking the glossary into `iSEEu:MarkdownBoard`. A bit more visible than having it tucked away into a dropdown, which tends to be a bit unwieldy...