Re: Publishing trees
Andrew Rambaut
It sounds like you want to construct a consensus tree (i.e., a tree that only
includes those clades with greater than a certain frequency in the posterior
sample). These are straight forward to construct using software such as PAUP
or Mega.
BEAST doesn't construct these for the simple reason that then you can't adequately
display temporal information like node heights/ages. A polytomy is constructed from
a number of nodes so it is not really meaningful to calculate an age.
We have take the approach of providing a representative tree that hopefully is
is in someway 'central' in the distribution of trees which you can use to display
the information and maintain the temporal dimension. You just have to make it
clear which nodes are well supported.
I am not sure why TreeAnnotator is not respecting your limit for annotating the
nodes but it is reasonably easy to remove low supported information in a graphics
editing package.
You can also use the 'Find' option in FigTree to highlight nodes with a posterior
support greater than (say) 80% and then colour them or annotate them with a symbol.
Best,
Andrew
On 29 Sep 2010, at 20:58, Rachel wrote:
> Thank you for responding, Derek and avalerio (since I have many times
> asked questions that were very important to me, that went unanswered).
>
> I should like to clarify- as I understand it, 'pruning' is the action
> of removing branches, or taxa, from my tree. I do not want to do this;
> instead, I want to collapse nodes with poor posterior support into a
> polytomy. I find that readers may draw the wrong conclusion when
> strict bifurcation is used in a phylogram, especially when polytomies
> are just as biologically plausible (if not more so).
>
> So far, I have tried FigTree, MEGA, Mesquite, and Dendrogram to create
> polytomies from my BEAST analyses, to no avail.
>
> Please tell me your thoughts/methods on creating/publishing
> polytomies.
>
> Rachel
>
> On Aug 23, 1:57 pm, avalerio <aavale...@gmail.com> wrote:
>> I have you try dendroscope??
>>
>> http://www-ab.informatik.uni-tuebingen.de/software/dendroscope/welcom...
>>
>>
>>
>> On Mon, Aug 23, 2010 at 2:59 PM, Derek Braun <cajunche...@gmail.com> wrote:
>>> If nobody else has a better answer, you should be able to write a
>>> script (php, Python) to prune the trees yourself. But that would take
>>> some time to write and debug. Definitely easier to see if there's a
>>> ready-made solution.
>>
>>> On Aug 4, 3:45 pm, Rachel <teatheded...@gmail.com> wrote:
>>>> Hallo all-
>>
>>>> I work with a large data set of viral sequences sampled over time, and
>>>> the trees I generate are quite large. I want the trees to be as clean
>>>> as they can for the readers, and in the past with non-Bayesian methods
>>>> I would collapse all nodes with bootstrap support below a specified
>>>> percentage. I have seen Bayesian phylograms published with no apparent
>>>> lower limit to the posterior probability (PP) (all nodes are
>>>> published, even those with low support) and I have talked with people
>>>> who would not publish nodes below a PP of 95%. The problem for me is
>>>> that when I use TreeAnnotator and FigTree, I haven't found a way to
>>>> collapse nodes below a specified PP threshold. I have tried exporting
>>>> a newick format tree from FigTree to use in another program like MEGA,
>>>> but the PP values do not go along. Is there another tree-drawing
>>>> program that can use the annotated nexus format that FigTree can
>>>> export?
>>
>>>> thanks!
>>>> Rachel
>
Andrew Rambaut
The BEAST manual is incorrect, it is the tree with the maximum PRODUCT of the posterior probabilities.
Leaving that aside, you are correct that TreeAnnotator finds a single tree. If the number of tips is small and/or the number of highly supported nodes is large then there may be multiple trees that have the same product, in which case TreeAnnotator picks an arbitrary one (probably the first it came across but I would need to check that). The point is that it then annotates that with the mean and intervals for the age of the nodes across all trees that have those nodes. It probably also makes sense to write the tree with branch lengths given by the mean (or median) across trees too (in fact we have just switched to that option being the default).
Andrew
On 30 Sep 2010, at 10:13, Nicola Bergh wrote:
> Dear BEAST users
> I have a query about the maximum clade credibility (MCC) tree often used as the 'target tree' in TreeAnnotator. Is this a single tree (from the 2007 BEAST user manual: "(it is) THE tree in the posterior sample that has the maximum SUM of posterior probabilities on its n-2 internal nodes")?
>
> My interpretation of this is that the TreeAnnotator determines, for every node recovered in at least one of the posterior sample of trees, the PP value (the percentage of trees in the sample that possess that node). It then finds the single tree in the sample that possesses as many as possible of these nodes. Does that sound correct? But it doesn't fully make sense to me: surely there will be many such trees, all with identical topology but with slightly different branch lengths? In which case, how does TreeAnnotator choose between them?
>
>
>
>
>
>
>
> Dr. Nicola Bergh
> The Compton Herbarium, Kirstenbosch Research Centre
> South African National Biodiversity Institute
> Private Bag X7, Newlands, Cape Town 7735, RSA
> Kirstenbosch National Botanical Garden, Rhodes Drive, Newlands, Cape Town, South Africa
> Tel +27(021)799-8800/8724 * Fax 27(021)761-4151 * www.sanbi.org * n.b...@sanbi.org.za
> --
> You received this message because you are subscribed to the Google Groups "beast-users" group.
> To post to this group, send email to beast...@googlegroups.com.
> To unsubscribe from this group, send email to beast-users...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/beast-users?hl=en.
>
>
>
>
> VISIT OUR NEW WEBSITE ----BIODIVERSITY ADVISOR---- ON HTTP://BIODIVERSITYADVISOR.SANBI.ORG/
>
> Think before you print. Please consider the environment before printing this email.
>
> NOTE: This e-mail message and any attachments are intended for the addressee only, and contain confidential information that may be legally privileged and/or the subject of copyright that is protected by law. Any unauthorised usage, disclosure, alteration or dissemination is prohibited. SANBI accepts no responsibility for loss, data corruption or mail that fails to reach its intended destination. Furthermore, SANBI cannot assure the integrity of this communication nor guarantee that it is free of errors, viruses, interception or interference. No liability, whether direct or indirect, is accepted by SANBI or the sender. Any view or opinion expressed in this message may not necessarily be that of SANBI or SANBI Management. SANBI reserves the right to monitor all e-mail communication.
>
> The disclaimer is located at http://www.sanbi.org/index.php?option=com_content&view=article&id=916:-sanbi-email-disclaimer&catid=1:about-sanb
Rachel
be able to move forward.