What do we want to call the x-axis on the tree viz?

[metasoarous]

Some options:

• Distance from naive
• Evolutionary time
• omit entirely

CC @matsen @lauradoepker ?

[lauradoepker]

What metric is being measured? Is it %mutation? Name it whatever the measurement is. Distance from naive (hamming distance?).

[metasoarous]

I think what we want here is: Evolutionary distance from naive

In our case (dnaml), this is certainly not % mutation or Hamming distance. If we were using parsimony (dnapars), it might be fair to think of it that way. But inherent in the ml of dnaml is the fact that we are looking at a maximum likelihood estimation of a probabilistic phylogenetic model. And in that model, if different mutations have different probabilities, then we'd expect the tree to reflect those probabilities in it's branch lengths. More concretely, if we see an uncommon mutation separating some cluster of sequences from the rest, we would most likely estimate the split to be more evolutionarily distant/long-ago than if it were a very common mutation.

For now, I think "Evolutionary distance from naive" most accurately reflects the model we're using, while also (hopefully?) coming off more intuitively and more generally than "Evolutionary time". Simply "Distance from naive" is in a sense more general if we were showing trees from other methods, since it implies it could be different distances depending on the tree method. As we get into new methods, perhaps it would be nice to have customization of these labels based on the build method metadata (which we already have). But for now I think something as clear as possible with respect to current data pipeline is thoughts. Additionally, with vega-lite export (see #52), you'd easily be able to export the plot to an editor where you could customize the labels.

Thoughts @lauradoepker ? Is this clearer to you?