-asr with codons

[Nathanael Walker-Hale]

Is it possible to do ancestral sequence reconstruction with codons yet? I tried specifying -st CODON and giving a codon substitution model, but it spat out a developer error, which makes me think you may not have implemented it yet.

Best wishes,

Nat Walker-Hale

[Minh Bui]

Hi Nat,

You are right, ASR feature does not work with codon models yet, which I overlooked. Thanks for pointing this out. We will fix it for the next 1.6.x release.

Minh

--
You received this message because you are subscribed to the Google Groups "IQ-TREE" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iqtree+un...@googlegroups.com.
To post to this group, send email to iqt...@googlegroups.com.
Visit this group at https://groups.google.com/group/iqtree.
For more options, visit https://groups.google.com/d/optout.

[Minh Bui]

Dear Nat,

Can you please tell me, if there is a numerical underflow message? Did you see have any NOTE about some state(s) not present in the alignment? If so, this is the problem and not specific to codon models. It would be helpful if you send me the log file as well.

Thanks,

Minh

[Nathanael Walker-Hale]

Dear Minh,

When I run it on e.g. the attached alignment, using the latest 1.7b release, I get the following error:

Do NNI search on 20 best initial trees
Estimate model parameters (epsilon = 0.100)
BETTER TREE FOUND at iteration 1: -81606.280
ERROR: iqtree.cpp:3206: virtual pair<int, int> IQTree::optimizeNNI(bool): Assertion `curScore > appliedNNIs.at(0).newloglh - 0.1 && "Using one NNI reduces LogL"' failed.
ERROR: STACK TRACE FOR DEBUGGING:
ERROR:
ERROR: *** IQ-TREE CRASHES WITH SIGNAL ABORTED
ERROR: *** For bug report please send to developers:
ERROR: ***    Log file: Bv_genbank_lab_DODA_4a_1_1.cds.fa.fsaaa2cdn.aln-cln.log
ERROR: ***    Alignment files (if possible)
Aborted (core dumped)

I have attached the alignment and the log.

Best,

Nat

[Minh Bui]

Dear Nat, 

I see, in this case I don’t recommend to use codon models. Your alignment has only 275 codon sites for 471 sequences, which cause numerical problem for such complex codon models. Please switch to DNA models only.

Cheers

Minh

<Bv_genbank_lab_DODA_4a_1_1.cds.fa.fsaaa2cdn.aln-cln><Bv_genbank_lab_DODA_4a_1_1.cds.fa.fsaaa2cdn.aln-cln.log>

[Nathanael Walker-Hale]

Hi Minh,

This was actually just a dataset I grabbed to reproduce the issue, but in general will codon reconstruction work? (in cases where n >> k)

Best,

Nat

[Minh Bui]

Yes, in general when you have enough data it should work.

Minh