Support ladderizing phylogenies and rotating branches

[gaurav]

Our current goal for curation is for the phylogeny to visually replicate the one in the paper as closely as possible, which can be done by tweaking the Newick string. To this end, it might be useful to allow the user to modify the phylogeny displayed in the Curation Tool in ways that do not alter the relationships of the node, such as by allowing branches to be rotated around a phylogeny or for the phylogeny to be ladderized.

[hlapp]

Rotating branches doesn't change the tree, so I'm not sure it's worth spending much time on that? I.e., achieving the goal of the phylogeny to match the one in the paper as closely as possible should include to match the topology exactly, and whether branches are rotated or not has no influence on that.

[GZhang2]

Ladderizing a tree greatly facilitates reading the tree, as successive sister clades can be seen in the same direction (top to bottom or left to right). Reproducing the reference tree topology the way it is printed in a publication is actually going to be tricky. When a digital tree file is read and the tree is drawn, it may or may not show the same "spatial topology" (order and orientation of tips as seen on a screen; not phylogenetic relationship) as the printed reference tree. I'd say say always ladderize a tree when it is drawn on the curation tool interface. I think this will facilitate human validation of clade resolution as it will enable easy reading of the tree, which will facilitate human-based reasoning and inference.

[gaurav]

Thanks for your feedback! I've modified the issue to clarify what I meant by saying "visually replicate the one in the paper as closely as possible". While the phylogeny should definitely have exactly the same topology as in the paper, I think it should also look as much like the figure in the paper as possible, which would make it easier to confirm that the phyloreferences are resolving as expected by the paper. I understand that this is going to be tricky, but I think it will make curation easier.

We could come up with a way to ladderize a phylogeny on the fly, but it would be much easier to ladderize a phylogeny by duplicating it, so we leave the existing phylogeny as-is and add a new phylogeny that rearranges the Newick string to ladderize it. So that might be a first step once we get around to implementing this issue.