Discrete reconstruction and transition rates

[adrian]

Dear Beast users,

I am performing a discrete trait reconstruction in Beast 1.10.4 and I have some issues identifying the rate of transition between trait states. Beast provides a log file exclusive for transition between trait states, but mean values are incredible high 0.7 to 1.5 for transitions/branch/million year. Can anyone advise my about the meaning of this values (maybe transitions/million years)?

Thank you in advance.

A.

[Philippe Lemey]

Hi Adrian,

The transition rates that are logged are not per branch and unit of time, but only per unit of time (so per million of years in your case). I personally would not draw much quantitative conclusions from them as they are difficult to inform by the data and hence they stay close to their prior expectation.

Best wishes,

Philippe

--
You received this message because you are subscribed to the Google Groups "beast-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beast-users...@googlegroups.com.
To post to this group, send email to beast...@googlegroups.com.
Visit this group at https://groups.google.com/group/beast-users.
To view this discussion on the web visit https://groups.google.com/d/msgid/beast-users/d65f83f5-0fc1-436f-801a-e253911648f9%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[adrian]

Dear Philippe,

thank you for your answer. My idea is to compare the pattern of transitions between different state, so probably with Beast information is enough.

Regards,

A.

El domingo, 2 de junio de 2019, 21:21:16 (UTC+2), philippe escribió:

Hi Adrian,

The transition rates that are logged are not per branch and unit of time, but only per unit of time (so per million of years in your case). I personally would not draw much quantitative conclusions from them as they are difficult to inform by the data and hence they stay close to their prior expectation.

Best wishes,

Philippe

On 2 Jun 2019, at 14:55, adrian <adrianvi...@gmail.com> wrote:

Dear Beast users,

I am performing a discrete trait reconstruction in Beast 1.10.4 and I have some issues identifying the rate of transition between trait states. Beast provides a log file exclusive for transition between trait states, but mean values are incredible high 0.7 to 1.5 for transitions/branch/million year. Can anyone advise my about the meaning of this values (maybe transitions/million years)?

Thank you in advance.

A.

--
You received this message because you are subscribed to the Google Groups "beast-users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to beast...@googlegroups.com.