mPTP

[blai...@gmail.com]

I am running mPTP on a new data set and am getting some odd results with the single rate method. ML analysis suggests 43 species. However, when I run a MCMC analysis I get a different value depending on what I look at. At the end of the MCMC run it says that the number of delimited species is 43, but the CCI and HPD intervals are (26, 30). The svg file also shows fewer than 43 species. Any idea what might be happening? Thanks.

Chris

[Alexandros Stamatakis]

Hi Chris,

What do you get when you run the multiple rate mode which we recommend?

Maybe you have many very short branches in the tree?

It would be helpgul if you could share the output.

Alexis

> --
> You received this message because you are subscribed to the Google
> Groups "PTP and GMYC species delimitation" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to ptp-species-delimi...@googlegroups.com
> <mailto:ptp-species-delimi...@googlegroups.com>.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/ptp-species-delimitation/850d800e-3407-43d7-9e83-ad4b5cf08023n%40googlegroups.com <https://groups.google.com/d/msgid/ptp-species-delimitation/850d800e-3407-43d7-9e83-ad4b5cf08023n%40googlegroups.com?utm_medium=email&utm_source=footer>.

--
Alexandros (Alexis) Stamatakis

ERA Chair, Institute of Computer Science, Foundation for Research and
Technology - Hellas
Research Group Leader, Heidelberg Institute for Theoretical Studies
Full Professor, Dept. of Informatics, Karlsruhe Institute of Technology

www.biocomp.gr (Crete lab)
www.exelixis-lab.org (Heidelberg lab)

[Paschalia Kapli]

Hi Chris,

The number of species printed on the screen after an MCMC run is the ML delimitation under the model used (single or multi). In the case you describe, it seems that the MCMC sampling does not support the ML delimitation under the single rate model. The ASV value should be quite low as well.

In general that's a sign that the model used does fit the data. I hope you get more consistent results under the multi rate model.

Best regards,

Paschalia

Στις Πέμ 20 Ιουλ 2023 στις 3:01 μ.μ., ο/η blai...@gmail.com <blai...@gmail.com> έγραψε:

I am running mPTP on a new data set and am getting some odd results with the single rate method. ML analysis suggests 43 species. However, when I run a MCMC analysis I get a different value depending on what I look at. At the end of the MCMC run it says that the number of delimited species is 43, but the CCI and HPD intervals are (26, 30). The svg file also shows fewer than 43 species. Any idea what might be happening? Thanks.

Chris

--

You received this message because you are subscribed to the Google Groups "PTP and GMYC species delimitation" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ptp-species-delimi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ptp-species-delimitation/850d800e-3407-43d7-9e83-ad4b5cf08023n%40googlegroups.com.

--

Paschalia Kapli, PhD

Telford Lab and Yang Lab

Department of Genetics, Evolution and Environment,
University College London,
Darwin Building,
Gower Street,
London WC1E 6BT, UK.

[blai...@gmail.com]

Hi Alexis and Paschalia, 

Thank you for your comments. I generally run both the single and multi rate methods to assess congruence. The multi rate analyses seemed to lump many "good" species together, consistent with some of my previous analyses suggesting that this method may be a bit conservative. The single rate method appeared more consistent with current taxonomy, but the MCMC results were conflicting with ML. It does seem like the data/tree may not be a good fit to the models. 

Chris

[Alexandros Stamatakis]

Dear Chris,

> Thank you for your comments. I generally run both the single and multi
> rate methods to assess congruence. The multi rate analyses seemed to
> lump many "good" species together, consistent with some of my previous
> analyses suggesting that this method may be a bit conservative. The
> single rate method appeared more consistent with current taxonomy, but
> the MCMC results were conflicting with ML. It does seem like the
> data/tree may not be a good fit to the models.

Yes, that might be the case, there are various things you can try:

1. Assess the impact of br-lens on the delimitation by generating BS
replicates and then re-optimzing the br-lens on the best scoring ML tree
(while keeping the tree topology fixed) and inputing these trees with
bootstrapped branch lengths into mPTP

2. Check how strong the tree signal is with our PYTHIA tool

https://academic.oup.com/mbe/article/39/12/msac254/6832260

and maybe consider using what we call a plausible tree set instead of a
single tree (if the signal in your dataset is weak) as input to mPTP,
i.e., delimit species on several uequally likely trees.

Alexis

> https://groups.google.com/d/msgid/ptp-species-delimitation/850d800e-3407-43d7-9e83-ad4b5cf08023n%40googlegroups.com <https://groups.google.com/d/msgid/ptp-species-delimitation/850d800e-3407-43d7-9e83-ad4b5cf08023n%40googlegroups.com?utm_medium=email&utm_source=footer>.

>
>
>
> --
> Paschalia Kapli, PhD
> Telford Lab and Yang Lab
> Department of Genetics, Evolution and Environment,
> University College London,
> Darwin Building,
> Gower Street,
> London WC1E 6BT, UK.
>

> --
> You received this message because you are subscribed to the Google
> Groups "PTP and GMYC species delimitation" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to ptp-species-delimi...@googlegroups.com

> <mailto:ptp-species-delimi...@googlegroups.com>.

> To view this discussion on the web visit

> https://groups.google.com/d/msgid/ptp-species-delimitation/b0873844-1eeb-4a1c-9af3-60f38952d688n%40googlegroups.com <https://groups.google.com/d/msgid/ptp-species-delimitation/b0873844-1eeb-4a1c-9af3-60f38952d688n%40googlegroups.com?utm_medium=email&utm_source=footer>.