Significant tree length change in TreeParser

58 views
Skip to first unread message

Diego Caraballo

unread,
Mar 18, 2025, 6:00:58 PMMar 18
to beast-users
Hi everyone,

I am running a speedemon species delimitation analysis in which I entered a starting newick tree, and used two calibrated nodes. Everything seems to work fine but I get the following warning message:

WARNING: Adjust tip heights attribute set to 'true' in TreeParser
         has resulted in significant (>0.1%) change in tree length.
         Use adjustTipHeights='false' to override this default.
  original max tip age =    0.003
       new max tip age =    0.000
  original tree length =    0.012
       new tree length =    0.021
       TL scale factor =    1.750
Log Likelihood Correction = -4090.7497913114744

I don't know if I should proceed with the run as if nothing is happening or if I should use adjustTipHeights='false', and what would be the effect of that modification in the XML.

Cheers,

Diego

Hee Han

unread,
Jun 9, 2025, 1:55:04 AMJun 9
to beast-users
Hi Diego,

Did you fix this problem now? I am also encountering the same error as you.
I hope to hear from you soon :)

Best,

Hee

2025년 3월 19일 수요일 오전 7시 0분 58초 UTC+9에 Diego Caraballo님이 작성:

higg...@gmail.com

unread,
Jun 9, 2025, 1:59:52 AMJun 9
to beast-users
Hi Diego, Hee,

If the Newick tree has branch lengths specified, you probably want to use adjustTipHeights='false'. This ensures that initial tip ages will be set according to the Newick tree.

Cheers, 

Remco

Hee Han

unread,
Jun 9, 2025, 4:45:41 PMJun 9
to beast-users
Hi Remco,

Thank you for your help! My newick tree just denotes its topology, so I change not only the status of the "adjustTipHeights" to false but also the volume of the "threshold" to larger. And it works normally! 
Thank you very much again for your effort.

Best,

Hee
2025년 6월 9일 월요일 오후 2시 59분 52초 UTC+9에 higg...@gmail.com님이 작성:

higg...@gmail.com

unread,
Jun 9, 2025, 4:47:25 PMJun 9
to beast-users
OK, just make sure to check that the tip dates in the trees that are sampled are correct.

Remco

Diego Caraballo

unread,
2:37 PM (3 hours ago) 2:37 PM
to beast-users
Thank you, Remco, that was exactly the problem. 
Reply all
Reply to author
Forward
0 new messages