Biostar Beta. Not for public use.
MODEL AVERAGING AND PHYLOGENETICS
0
Entering edit mode
5.0 years ago
nitishd.dave • 20
@nitishd.dave22277

I am using jMODELtest2 for model averaging however the given model is not compatible with the BEAST software. Can anyone suggest how use the given model in BEAST as it equipped with only three models?

How to use a model for Phylogenetic analysis generated by jMODELtest2 in BEAST software ?

myposts • 860 views
ADD COMMENTlink
0
Entering edit mode

Can you give more info. What is the suggested model ? Usually you would go to the next best alternative. So move down the list until you find one compatible with BEAST. However, be aware your subsequent model is the best out of only the models beast runs and not the best of all models in JmodelTest.

ADD REPLYlink
0
Entering edit mode

I do not think that it will work. I do not know BEAST, but I know that there are many other phylogenetic software that do not implement all model of evolution. I guess either you restrict you model prediction or you dig a bit more among the phylogenetic software to see which one support the model that better applies to your needs.

ADD REPLYlink
1
Entering edit mode

You absolutely can implement other models in BEAST. See my answer below.

ADD REPLYlink
0
Entering edit mode

Also, This may help if you want to make a robust phylogeny : http://www.mdpi.com/2079-3197/3/2/177

ADD REPLYlink
1
Entering edit mode
5.0 years ago
Brice Sarver ♦ 2.6k
@Brice Sarver13598

Other models of nucleotide sequence evolution can be manually implemented by adding to the XML file produced by BEAUTi. Start here and look for other resources by Joseph Brown - he had an example on the BEAST page pre-2.0. However, if your model is a subset of the GTR model, you might consider selecting GTR and letting the parameters be estimated that way. Because priors are placed on them, any rate class not supported by your data ought to tend toward zero. I think Alexei Drummond argues for this under a Bayesian philosophy in the user group or somewhere - I can't remember offhand.

ADD COMMENTlink
0
Entering edit mode

Thank you brice, i want to ask one more question that, i tried to build a molecular clock of particular scorpion specimen but i dont know whether i should trust it or not. I sequenced coi region as well as 16s region. of which 16s r rna region of mitochondria is thought to be evolve more slowly and conservatively. i built it on beast with gtr + g + i model. how can i confirm that the given molecular clock is correct.

ADD REPLYlink
0
Entering edit mode

Your BEAST tree will be clocklike, but will be in relative units of time after estimating among-lineage scalars for the branches. In order to convert to absolute time, you'll need to place prior distributions on node ages or scale the tree by time after the fact. The first way is the most common way to get an ulltrametric tree. There are tutorials on how to do this online, but basically you can place a prior on the node depth after constraining it to exist in your starting tree (else the likelihood would be zero!).

ADD REPLYlink
0
Entering edit mode

Thanks again brice, can i simply put the fossil date calibration point on the prior distribution. and i am using only one calibration point for as i have a fossil record of that specimen at family level. will adding one more calibration point will be of any help ?

ADD REPLYlink
0
Entering edit mode

Yes. The more, the better.

ADD REPLYlink

Login before adding your answer.

Similar Posts
Loading Similar Posts
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.3