codon models

[alex]

Hello and happy new year!

I would like to ask about substitution mapping. I tried to use the M3, M7 and M8 models for the DN and DS counts but the numbers I get are equal to zero. Are they not implemented for mapping or I am doing something wrong? If those models are not implemented for substitution mapping, which models are? Only the M0 model seems to do the counting.

Alex

[Julien Yann Dutheil]

Dear Alex,

Happy new year to you as well!

Indeed, substitution mapping does not currently work with mixed models
:s We are currently re-implementing the likelihood classes in a more
generic framework to allow this type of analysis. I will disable the
use of mixed models in mapnh for now, sorry for that.

All the best,

Julien.

> --
>
>
>



--
Julien Y. Dutheil, Ph-D
0 (+49) 6421 178 986

§ Max Planck Institute for Terrestrial Microbiology
Department of Organismic Interactions
Marburg -- GERMANY

§ Intitute of Evolutionary Sciences - Montpellier
University of Montpellier 2 -- FRANCE

[alex]

a ok!! :P Thank you very much!

[alex]

Dear Julien,

I would like to ask you about the F3X4 frequencies for codons. I am using YN98 model and it asks for 9 freq_Codon1_Full.theta values. What are those parameters are and how could I define them? Thanks in advance!

Yours sincerely,
Alex

[Julien Yann Dutheil]

Dear Alex,

I assume you are using MapNH? You can provide manual values for these
parameters, but the best way is actually to estimate them. In order to
do that, you can use bppml, as we did in the examples provided with
the source code.

Hope this helps,

Julien.

[alex]

Dear Julies,

Thanks! Yes, I use MapNH. I try to use useObservedFreqs=yes as in the examples, but the program suggests instead to use the parameter initFreqs=observed. Is that for the estimation of the theta codon values?

Yours sincerely,
Alexandra