HELP-continuous phylogeography

Vaca

unread,

May 23, 2022, 3:25:42 PM5/23/22

to beast-users

Dear all,

I have one problem regarding continuous phylogeography.

KML file shows root in one country. However, mcc tree shows root in other country. Country which was set as a root in KML file is located in the basal part of the mcc tree, but it is not in the root.

Why this is happening?

thanks

Vaca

Pavel Rinkman

unread,

May 27, 2022, 10:37:06 AM5/27/22

to beast-users

Do you use GEO_SPHERE from BEAST 2 package set or continuous phylogeography on plain from the first one?

понедельник, 23 мая 2022 г. в 22:25:42 UTC+3, Vaca:

Vaca

unread,

May 29, 2022, 6:56:47 PM5/29/22

to beast-users

I used BEAST 1.10.4.

Junshi Dong

unread,

May 29, 2022, 7:09:08 PM5/29/22

to beast-users

Hello Vaca,

I couldn't reconstruct the phylogeny in continuous space. It showed an error message. Do you have any idea why this is? Thank you.

Junshi

Pavel Rinkman

unread,

May 30, 2022, 4:42:29 AM5/30/22

to beast-users

Hello, Vaca!

Could you provide a link to your xml?

понедельник, 30 мая 2022 г. в 02:09:08 UTC+3, Junshi Dong:

Pavel Rinkman

unread,

May 30, 2022, 4:49:47 AM5/30/22

to beast-users

Hello, Junshi!

Could you give a link to your tree? SpreaD3 has a lot of little problems like this, for example, it is broken at specific dates in a tree (DST change days). So it would be very useful to look at your tree to define the problem in your case.

понедельник, 30 мая 2022 г. в 02:09:08 UTC+3, Junshi Dong:

Hello Vaca,

Junshi Dong

unread,

May 30, 2022, 2:52:57 PM5/30/22

to beast-users

Hello Pavel,

Thanks for the reply! Actually, when I did the discrete phylogeographical diffusion, I had problems with spreadD3 as well so I switched to a previous version of Spread and an online spread tool (https://spreadviz.org/home), which both seem to work just fine. But it ran into some issues again when I tried to calculate the Bayes factor. For the continuous one, here is the link to the zip file that contains the tree and the log file Beast generated, and the XML file ( https://drive.google.com/file/d/1NErA_bUHSerPeFDjrl0gNYMw75a7KjY6/view?usp=sharing). Also is there any other way to talk to you? Since I don't belong to this group so my message will always get sent out delayed. Thanks a lot again!

Sncerely,

Junshi

HS

unread,

May 31, 2022, 4:47:33 AM5/31/22

to beast-users

Hi,

Do you also annotate the trees file with Treeannotator (available in the Beast package)? The resulting MCC tree has to be uploaded to SpreaD3, not the complete trees file. Although, there is a way also to do with the whole trees file (better a thinned subset), but it is probably with a different setting or even a program, don't remember well.