Dear Ivan.
phytools::fitMk does not do ancestral state reconstruction. What
function did you use to obtain the estimates at internal nodes?
If you want to constrain the states at particular nodes during ancestral
state reconstruction this can be done in a fairly straightforward way.
Here is a demo that I posted to my blog some years ago:
http://blog.phytools.org/2014/06/constraining-internal-node-values-using.html.
The demo is for stochastic mapping using make.simmap but it would work
in a similar way using ape::ace or other methods, and can also be used
for simple model fitting with fitMk or geiger::fitDiscrete.
All the best, Liam
Liam J. Revell
Associate Professor, University of Massachusetts Boston
Profesor Asistente, Universidad Católica de la Ssma Concepción
web: http://faculty.umb.edu/liam.revell/, http://www.phytools.org
Academic Director UMass Boston Chile Abroad (starting 2019):
https://www.umb.edu/academics/caps/international/biology_chile
On 9/18/2019 4:36 PM, Ivan Magalhaes wrote:
> [EXTERNAL SENDER]
>
> Hello everyone,
> Please, I am trying to use the fitMk function in phytools to test different
> scenarios for character evolution (equal rates, all-rates-different, and
> customized matrices with irreversible states) and getting some unexpected
> results.
> Specifically, I have a character that has a defined ancestral condition
> (established by outgroups not present in the analysis), three different
> derived conditions, and taxa which lack this particular structure and have
> been coded as "absent". A small group of taxa which is deeply nested in the
> tree presents the plesiomorphic condition, and I would like to test if this
> is a plesiomorphic retention or a re-gain. In the all-rates-different model,
> it is optimized as a re-gain. So I wanted to compare this with a model where
> re-gains are impossible, using this custom model:
>
> - "ancestral" can shift to any state-"derived" states can shift to "absent"
> only, but not to other derived states, nor reverse to the "ancestral"
> condition-"absent" cannot shift to any condition
>
>
> This gives me a tree in which the character is optimized as a plesiomorphy.
> Everything seems fine so far!
> However, the exact same optimization is produced if I use a different, less
> restrictive model where all transitions are possible except for re-gains from
> absent structures (i.e. the model allows for reversals from the derived
> states to the ancestral state, and also transitions among derived states). I
> find this very strange, because the only difference between this and the ARD
> model is the impossibility to reverse from "absent" to other states - but
> neither of the trees suggested such an optimization would occur. Thus, I
> would expect that this second customized matrix would give results identical
> to those of ARD -- please, does anyone know why such a result is to be
> expected?
> Many thanks in advance.
> Best regards,Ivan
>
> --
> Dr. Ivan L. F. Magalhaes
>
> División Aracnología
> Museo Argentino de Ciencias Naturales "Bernardino Rivadavia"
> Av. Ángel Gallardo 470, C1405DJR, Buenos Aires, Argentinatel +54 11 4982-8370
> int. 169
> researchgate.net/profile/Ivan_MagalhaesTwitter: @6eyes8legs
>
> [[alternative HTML version deleted]]
>
> ___
> R-sig-phylo mailing list - R-sig-phylo@r-project.org
> https://nam01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fstat.ethz.ch%2Fmailman%2Flistinfo%2Fr-sig-phylodata=02%7C01%7Cliam.revell%40umb.edu%7C6dd2604ffd2646aefcf608d73c6fa1ad%7Cb97188711ee94425953c1ace1373eb38%7C0%7C1%7C637044322481538463sdata=%2BiTS1Y1A7al96v%2Fnz7ckLVOgD1sMhNvCUq4BROMzf4g%3Dreserved=0
> Searchable archive at
> https://nam01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.mail-archive.com%2Fr-sig-phylo%40r-project.org%2Fdata=02%7C01%7Cliam.revell%40umb.edu%7C6dd2604ffd2646aefcf608d73c6fa1ad%7Cb97188711ee94425953c1ace1373eb38%7C0%7C1%7C637044322481538463sdata=Hi5V324VpX3gzfIyz9r5ULFE29y6olfpcFJX9%2B%2FV2as%3Dreserved=0
>
___
R-sig-phylo mailing list - R-sig-phylo@r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
Searchable archive at http://www.mail-archive.com/r-sig-phylo@r-project.org/
