I was wondering if there´s a possibility to represent the normalized values of only one transcript across several samples as a heatmap, Some kind of "ratio" heatmap that would tell me how "close" or "far" are all those samples from each other based only on the normalized values of transcript "A".
If not possible in a heatmap, do you know another way to represent it graphically?
Please note, I need to represent some kind of similarity, not correlation (since I have only one variable).
Thank you in advance!
Yeah, I already did that. (Easy, right?) :)
But I was wondering if there is a way to show (for instance with color intensity) how "alike" these samples are, besides a classical barplot or a scatterplot.
Just set thresholds based on the mean expression level or something to that effect and supply that as the coloring condition?
An addition to that could be a single axis plot with the centre being mean expression of your transcript of interest.
Maybe that could work, yes! Thank you.
Dunois option is valid and not too difficult to implement in ggplot. In general, though, I find that colour is a lot less expressive than the usual distances between points, bars or lines to represent quantitative variables. If at all possible I try to avoid heatmaps, which are effectively 3D plots with colour being the third axis.