Errors adding tree

[Czirion]

Hi, I'm having errors when trying to use ntSynt-viz with the --tree option

When I add a tree with branchlengths I get an error in the rule ribbon_plot_tree:

Error: Error: Tree could not be reordered within the maximum iterations.

When I add a tree without branchlengths I get an error in the rule cladogram:

Error in p$edge.length[which(p$edge[, 2] == dd)] <- p$edge.length[which(p$edge[,  : 
  replacement has length zero
Calls: <Anonymous> -> reroot
In addition: Warning message:
In dist.nodes(x) : the tree has no branch length: fixing them to one.

When I add the tree _est-distances.nwk created in a previous run of ntSynt-viz it works fine and displays the tree in the plot. But it does not have the true topology that I want to display.

Is there a way to provide a tree without branchlengths that would work?

Thank you

[lcoombe]

Hi @Czirion,

What version of ntSynt-viz are you using? And how many input genomes are in your tree?
I had seen a similar issue with 3 genomes in particular, but should have been fixed in the newest version.. but possible there are other edge cases that could cause issues.

If you feel comfortable, feel free to share your input files and I could take a look - either here or by (lcoombe@bccrc.ca)

Thank you for your interest in ntSynt-viz!
Lauren

[Czirion]

Hi Lauren,

I am using ntSynt-viz v1.0.2. And I'm having this issue with different number and combination of genomes. I can't share the genomes that I am working with, but I'll see if I get the same issue with public genomes and I'll let you know.

Thank you for developing ntSynt-viz and for your help!
Claudia

[lcoombe]

Thanks for the additional info, and totally understandable.
So I could try to reproduce on my side, what range of numbers of genomes have you tried? Do you know if you only see it when you input trees without branch lengths, or if you see it with and without branch lengths?

[lcoombe]

Hi Claudia,

Thank you again for reporting this issue and helping me to reproduce it!

I have implemented fixes in #33 (now merged to main branch):

1. If an input tree does not have branch lengths, the branch lengths are set to 1 automatically
2. If the input tree is already rooted, do not midpoint root it (this was being done erroneously before, causing the issue you saw)

Feel free to test it out on your end (from the main branch code) - I will be releasing a new ntSynt-viz version with these fixes tomorrow.

[Czirion]

Hi Lauren,

It works now! Thank you so much for solving this!

PS. I noticed that if the target genome is in a polytomy, its clade is put on top, but the target genome itself is not. I don't think this will be a common situation, but just so you know.

Thank you!

[lcoombe]

Hi Claudia,

Awesome, I'm so glad it works on your end now! I'll release a new version of ntSynt-viz with that fix shortly.

Thanks for letting me know about the polytomy issue - I hadn't tested ntSynt-viz in that situation yet. I'll see if I can find some data to reproduce that on my side as well and see if there is a fix I could do!