Resample at lower frequency before creating MCC tree?

[Peter]

Hi,

I have a question regarding resampling/thinning in LogCombiner and creating MCC trees in TreeAnnotator. For a publication I would like to present both a posterior distibution of trees (e.g. 1000 or 10 000 trees) as well as an MCC tree and I was wondering whether these should be generated separately or whether it is fine to first combine trees and resample at a lower frequency in LogCombiner, and then generate an MCC tree from the resulting distribution of trees using TreeAnnotator?

I have been running four independent chains in BEAST v1.8 for 250 million generations each, sampling every 5000 generations. If I was to combine these in LogCombiner (after removing 50 mill. generations as burn-in from each chain) without resampling at a lower frequency, I would end up with 160 000 trees. 

Thus, is it fine to resample these trees at a lower frequency that would give me say 1000 trees and create an MCC tree from those? Or should I ideally input all 160 000 trees into TreeAnnotator?   

Cheers,

Peter

[Santiago Sánchez]

Hi Peter,

Depending on how large your trees are, it is probable that TreeAnnotator will crash with that much trees (or run out of memory). What you could do is combine your trees and then resample to end up having ~10,000 trees. You could do this with a perl script I wrote called "GetPosteriorTrees.pl". You can find it in: https://sites.google.com/site/santiagosnchezrmirez/home/software/perl under "Trees". If you run it with the -h argument you'll have the details on how it works. If you have questions, let me know.

Note that it will only work with trees generated in BEAST 1.x, not BEAST 2.x.

Cheers,

Santiago

--
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.
For more options, visit https://groups.google.com/d/optout.

--

Santiago Sánchez-Ramírez

Environmental Genomics Group

Max Planck Institute for Evolutionary Biology

August-Thienemann-Str. 2
24306 Plön
Germany

Website: https://sites.google.com/site/santiagosnchezrmirez/