how to find TMRCA, bayes factor using BEAST

Question

[rajeshkumar_vinod]

Hello everyone,

I have to find TMRCA, bayes factor using BEAST for given set of sequences and models.

as far as i could understand is for TMRCA we will have to go to taxa tab in beauti, create taxon set then generate XML file and run it in BEAST?

for bayes factor i assume that we will have to go to MCMC tab in beauti and choose a method to calculate marginal likelihood estimation(MLE) from MLE how can we calculate bayes factor?

Thanks.

Answer 1

[Prakki Rama]

1) To calculate MLE:

--> In Beauti, Go to Manage Packages -> Select Model Selection and Install Model Selection.

--> Run the appstore command from the beast package -> This should open window

--> From Filter, Select Model Selection -> Select PathSampler -> Load the XML file you created using BEAUTI -> Generates MLE for the specific model

Repeat this for other models

2) To calculate Bayes Factor

--> List down all the MLE for each model and select the model with highest MLE.

--> From the tutorial: the marginal likelihood estimate value for model 1 is X and the value for model 2 is Y, then the Bayes factor comparing model s1 and 2 is is X-Y. Simply to put, BF = (MODEL1 MLE - MODEL2 MLE)

3) To find tMRCA

--> Load the output from treeannotater into figtree

--> In the Time scale dropdown, under Scale by factor :

Offset by: Type the youngest tip (year)

scale factor: -1.0

--> In the Scale Axis dropdown, select Reverse axis, Automatic scale

--> Select Node labels

The left most node label is the Most Recent Common Ancestor (tMRCA)

Hope this helps!