Re: garli tree output format

[Derrick Zwickl]

Hi Brigida,

My guess is that the RAxML distance function is expecting fully bifurcating trees.  RAxML doesn't output trees with polytomies, but GARLI does (by default) collapse edges when the ML edge length estimate is zero.  That is the right thing to do, in my opinion, since otherwise you get trees with effectively random resolutions of polytomies.  Incidentally, collapsing those edges will probably make your gene trees more similar (if using an RF distance that ignores edge lengths), since trees may have different resolutions of zero length edges.

Let me know if you have other questions.

Best,
Derrick

On 2/16/15 10:38 AM, Brigida Gallone wrote:

Hello Derrik,

I used Garli to compute ML trees  for a big collection of genes and now I want to estimate the the normalized Robinson–Foulds tree distance, I always use RAxML to do this, but apparently there is incompatibility with the file format of the trees generated by Garli and the input trees ( regular newik format ) required by RaxML. I know that there are many ways out there to compute this distance but I am curios if I can or cannot indeed use the Garli output with RAxML. I cannot spot any obvious difference. The same trees work perfectly fine in Treedist or Clann by the way.

Many thanks for your answer.

Best,

Brigida

--
You received this message because you are subscribed to the Google Groups "garli_users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to garli_users...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.