Time scaled tree with IQTree giving off looking branch length

Question

[SushiRoll]

Hi everyone!

I'm having some weird results using IQTree to build a time-scaled tree. I have used the tool before but this is my first time doing this type of analysis. I have a set of 80 E. coli which I have submitted to an SNP analysis. I have now a core SNP alignment that I have run through Gubbins to remove recombination sites. This alignment I fed it to IQTree and used the time reconstruction like this:

iqtree -s alignment.fasta --date date.tsv

Where date.tsv is a two-column file with sample name and isolation date (Between 2012 and 2020).

I took the generated tree to IToL and plotted the branch lengths as time and this is (a fragment of) what came out:

As you can see the dates seem to be extremely large, plus shorter branches have a longer time distance. I wonder if I'm misinterpreting the whole thing or if doing an SNP alignment is simply too much since it will find too many differences.

Any help would be greatly welcome.

Answer 1

[cfos4698]

Did you remove invariant (constant) sites from the SNP alignment? If so, you should probably also be specifying an ascertainment bias correction. See: http://www.iqtree.org/doc/iqtree-doc.pdf#page=158.20

An ascertainment bias correction (+ASC) model (Lewis, 2001) should be applied if the
alignment does not contain constant sites (such as morphological or SNPs data). 

For example:
• MK+ASC: For morphological data.
• GTR+ASC: For SNPs data.
+ASC will correct the likelihood conditioned on variable sites. Without +ASC, the branch
lengths might be overestimated.

[SushiRoll]

Hey cfos! Thanks for the reply! I did apply the ascertainment correction model, in fact I applied GTR+F+ASC for this tree. I'm wondering if it has to do with the way I'm visualising the tree ...