*BEAST to estimate mutation rates

[Carla Hurt]

Hi All,

   I have a multilocus data set generated using GBS and I would like to use it to estimate the mean mutation rate.  I have 2-3 individuals per species and known divergence times.  I have done this using BEAST but I would like to use a coalescent based method that accounts for ancestral lineage sorting.  Is it possible to set up a *BEAST run to estimate the mutation rate parameter?  

[Santiago Sanchez]

Hi Carla,

Doing this with standard BEAST seems relatively easy if you use a concatenation approach. But in *BEAST you need to provide unlinked partitions for each of your loci. I'm not sure how practical or informative it would be to do this on GBS data. How many and how long are your loci? Maybe you could start by setting up *BEAST with unlinked gene trees for each loci (required), and have the substitution and clock models linked, just reduce complexity (and probably avoid convergence issues).

Just an opinion.

Cheers,

Santiago

Santiago Sanchez-Ramirez

Ecology and Evolutionary Biology, University of Toronto

Natural History (Mycology), Royal Ontario Museum

100 Queen's Park

Toronto, ON

M5S 2C6

Canada

+1 416 586 8025

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

On Aug 18, 2015, at 10:39 PM, Carla Hurt <hurtc...@gmail.com> wrote:

Hi All,

   I have a multilocus data set generated using GBS and I would like to use it to estimate the mean mutation rate.  I have 2-3 individuals per species and known divergence times.  I have done this using BEAST but I would like to use a coalescent based method that accounts for ancestral lineage sorting.  Is it possible to set up a *BEAST run to estimate the mutation rate parameter?  

--
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 http://groups.google.com/group/beast-users.
For more options, visit https://groups.google.com/d/optout.

[Carla Hurt]

Thank you for your insights.   Can you link the trees in starbeast?  My understading is limited but I thought that the assumption was that each gene had its own tree with a shared underlying species tree?

--
You received this message because you are subscribed to a topic in the Google Groups "beast-users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/beast-users/Sm7_8hivJ1w/unsubscribe.
To unsubscribe from this group and all its topics, send an email to beast-users...@googlegroups.com.

[Santiago Sánchez]

Gene trees cannot be linked in *BEAST. They are estimated separately but the multi-species coalescent model integrates them to infer species trees.

Santiago Sánchez-Ramírez

Department of Ecology and Evolutionary Biology, University of Toronto

Department of Natural History (Mycology), Royal Ontario Museum

100 Queen's Park

Toronto, ON

M5S 2C6

Canada

Other email: santiago...@mail.utoronto.ca

Tel. 416-586-8025

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

[Carla Hurt]

That makes sense.  Thank you so much for your help!

[Chris Martin]

Along these same lines,

for the same RADseq dataset, is there a difference in the estimated mutation rate between *BEAST and a concatenated BEAST analysis? 

I know that divergence times are affected, but...can't find any literature/examples of the effect on mutation rate of these two analyses? 

[Alexei Drummond]

*BEAST is probably more accurate if there is incomplete lineage sorting ;)

Which direction concatenation is biased in would depend on the method of calibration. If you calibrate an internal node on a concatenated analysis then I would guess that the mutation rate would usually be over-estimated since the amount of molecular divergence you will attribute to the time from speciation to the present would include ancestral polymorphisms so it will appear that more mutations have happened since the speciation then actually occurred.

Cheers

Alexei

Visit this group at https://groups.google.com/group/beast-users.

[Huw A. Ogilvie]

Concatenation biases are even worse when population sizes vary across a tree, because one source of bias (average time to coalescence) is dependent on population size. Also, the species tree topology is another source of bias (SPILS - see Mendes & Hahn 2016, as well as Mendes, Hahn & Hahn 2016).

- Huw

[Chris Martin]

Thank you both so much for the helpful explanation and useful references!