Dollo-like substituion model for binary characters

[Fritjof La]

Hi all,

I have a presence absence matrix of retroposon insertion data (0, 1 encoded). For retroposon insertion, typically only the gain of a character is possible (0 -> 1) because there is no molecular mechanism to remove a insertion. I think it would be super interesting to analyse this data using the quartet likelihood mapping implemented in IQ-TREE. However, to my understanding, the implemented substitution models for binary data, don't capture the evolutionary mechanism properly. I was wondering it providing a own matrix for that is possible or if you have some other advice how such data can be analyzed with IQ-TREE.

Thanks a lot

Fritjof

[Minh Bui]

Hi Fritjof,

I’m not sure if this is a valid phylogenetic Markov model or not. As the time goes by, you will only observe 1s because all 0s will become 1s. Whereas in a typical model, you can go from any one state to any other state either directly or indirectly (connectedness). However, your model is not connected…

Cheers

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.

[Fritjof La]

Hi Minh,

thanks. That makes a lot of sense! Would it be possible to have a connected model with a really low rate for 1 -> 0 changes and define this as model within IQTREE?

Best

Fritjof

[Minh Bui]

Hi Fritjof,

Sorry for the slow response. I looked at such model and there is numerical issue with the model. So setting a low rate is still problematic. My conclusion right now is that the model does not work.

Cheers

Minh