checking estimate box of clock rate panel in BEAUTI

103 views
Skip to first unread message

Sahar Javaheri

unread,
Nov 15, 2019, 10:58:21 PM11/15/19
to beast-users

Hi all,


I am trying to use Beast v2. 5. 1 for estimating divergence times. my problem is here, when I unchecked the estimate box in clock rate panel in Beauti. I can obtain convergence for some priors by Tracer but, when I check the estimate box, the result is different and I can’t obtain the convergence with any other priors as you can see in the pictures below,

image.png

Yule prior and HK model, check the estimate box:

image.png

Yule prior and HK model, uncheck the estimate box:

image.png


If I unchecked the estimate box of clock rate panel for obtaining the convergence will it be correct?


I really appreciate if you guide me about this problem!


Cheers,

Sahar

Remco Bouckaert

unread,
Nov 18, 2019, 2:48:14 PM11/18/19
to beast...@googlegroups.com
Hi Sahar,

The images you posted did not come through, so I am not completely sure what the problem is, but in general checking estimate boxes by hand should not be done lightly — especially for the clock rate. BEAUti automatically sets the clock rate estimate flag if there is some timing information, like tip dates or timing calibrations on clades. If there is no such timing information it is not possible to estimate the clock rate, and forcing it to be estimated results in failure to converge (unless there is a strong prior on the clock rate).

Cheers,

Remco

--
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 view this discussion on the web visit https://groups.google.com/d/msgid/beast-users/6a675333-473a-404d-9eae-7ae49b39fee8%40googlegroups.com.

Sahar Javaheri

unread,
Nov 26, 2019, 4:32:39 PM11/26/19
to beast...@googlegroups.com
Dear  Remco,
Thank you so much for your help. I have only the substitution rate of my target gene. Nevertheless, can BEAUTI automatically set the clock rate even if I unchecked the estimate box of clock rate panel? 

Regards,
Sahar



--


***************************************

Sahar Javaheri Tehrani
PhD student of Animal Biosystematics 
Department of Biology, Ferdowsi University of Mashhad, Mashhad, Iran

Remco Bouckaert

unread,
Nov 27, 2019, 1:44:28 PM11/27/19
to beast...@googlegroups.com
HI Sahar,

When you uncheck the “Mode/Automatic set clock rate” box, BEAUti will not attempt to set estimate boxes automatically, and you have to do this by hand.

If you have a prior on the clock rate, that is a case where you have to uncheck that box, check the estimate box for the clock rate (in the Clock Model panel), and define the clock prior (in the Priors panel). It is possible that if the clock rate prior is too wide (that is, there is little information about the clock rate) that the clock prior is not sufficiently informative to guide the analysis, and the chain does not converge because the clock rate and tree height are in practice not identifiable together.

Cheers,

Remco


On 27/11/2019, at 10:31 AM, Sahar Javaheri <sahar.ja...@gmail.com> wrote:

Dear  Remco,
Thank you so much for your help. I have only the substitution rate of my target gene. Nevertheless, can BEAUTI automatically set the clock rate even if I unchecked the estimate box of clock rate panel? 

Regards,
Sahar

On Mon, Nov 18, 2019 at 11:18 PM Remco Bouckaert <higg...@gmail.com> wrote:
Hi Sahar,

The images you posted did not come through, so I am not completely sure what the problem is, but in general checking estimate boxes by hand should not be done lightly — especially for the clock rate. BEAUti automatically sets the clock rate estimate flag if there is some timing information, like tip dates or timing calibrations on clades. If there is no such timing information it is not possible to estimate the clock rate, and forcing it to be estimated results in failure to converge (unless there is a strong prior on the clock rate).

Cheers,

Remco
On 16/11/2019, at 4:03 AM, Sahar Javaheri <sahar.ja...@gmail.com> wrote:

Hi all,


I am trying to use Beast v2. 5. 1 for estimating divergence times. my problem is here, when I unchecked the estimate box in clock rate panel in Beauti. I can obtain convergence for some priors by Tracer but, when I check the estimate box, the result is different and I can’t obtain the convergence with any other priors as you can see in the pictures below,

<image.png>

Yule prior and HK model, check the estimate box:

<image.png>

Yule prior and HK model, uncheck the estimate box:

<image.png>


If I unchecked the estimate box of clock rate panel for obtaining the convergence will it be correct?


I really appreciate if you guide me about this problem!


Cheers,

Sahar


-- 
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 view this discussion on the web visit https://groups.google.com/d/msgid/beast-users/6a675333-473a-404d-9eae-7ae49b39fee8%40googlegroups.com.

-- 
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 view this discussion on the web visit https://groups.google.com/d/msgid/beast-users/DB57C4C9-8B90-42B1-AAD4-3DD1D28CEBEB%40gmail.com.


-- 


***************************************


Sahar Javaheri Tehrani
PhD student of Animal Biosystematics 
Department of Biology, Ferdowsi University of Mashhad, Mashhad, Iran


-- 
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.

Sahar Javaheri

unread,
Dec 23, 2019, 7:05:32 AM12/23/19
to beast...@googlegroups.com

Dear Remco,

 

Thank you so much for your helpful information. Since, I want to estimate the node ages on the different subspecies groups in the species level by one gene and I have just substitution rate can I follow the below settings in Beast v.2.5.1:

Mode ….> check “Automatic set fix mean substitution rate flagand uncheck the “Automatic set clock rate” box

In clock model panel ….> insert the substitution rate and uncheck the estimate box.

Is it correct?

Thanks in advance for your valuable guides,

Sahar



Reply all
Reply to author
Forward
0 new messages