bModelTest is a BEAST 2 package, so you need to get BEAST 2 first.
To run a bModelTest model analysis, you need the bModelTest package. The easiest way to do this is by starting BEAUti, and open the package manager using the File/Mange packages menu. A dialog pops up that looks like this:
Select the bModelTest package and click the Install/Upgrade button. Restart BEAUti in order for the bModelTest model to be available.
Setting up an analysis in BEAUti
Start BEAUti, load an alignment as per usual. Now, when you go to the Site model panel, there is a choice of site models shown in the combobox at the top of the panel:
Select the BEAST Model Test model for the bModelTest model.
To analyse the results of a BEAST run with bModelTest, you can start Tracer and find the following items:
BMT_ModelInidicator is the index of the substitution model as listed in the appendix of the paper.
substmodel is the model represented as a 6-digit number, where the position of the digit refers to rates ac, ag, at, cg, ct and gt respectively, and equal digits indicates that rates are shared, so 111111 is Jukes Cantor (if frequencies are kept equal), 121121 is HKY, 123456 is GTR etc.
rateAC … rateGT are the rates according to the model. ESSs should be good for these rates, but if you plot joint-marginals of pairs you may find high correlation between some of these rates.
BMT_Rates.1 to 6 are the rates used to build up the rate matrix. If only low parameter models are samples, the higher rates will be sampled very infrequently, and you should expect low ESSs for them. Correlation between pairs of rates should be low.
BMT_gammaShape is the gamma shape parameter as it is being sampled. For parts of the chain that gamma rate heterogeneity is switched off, the parameter will not be sampled, and the trace will show periods where the parameter is stuck.
hasGammaRates indicates whether gamma rate heterogeneity it used (1) or not used (0). The mean can be interpreted as the proportion of time that gamma rate heterogeneity is switched on.
ActiveGammaShape is the gamma shape parameter when it is sampled, but it is zero when it it not sampled. To get the estimate of the mean of the shape parameter, divide the mean ActiveGammaShape by the mean of hasGammaRates.
BMT_ProportionInvariant, hasInvariantSites and ActivePropInvariant are the value for proportion invariable similar to BMT_gammaShape, hasGammaRates and ActiveGammaShape respectively.
hasEqualFreqs indicates whether equal frequencies are used and the mean can be interpreted as the proportion of time that equal frequencies is used.
To visualise the distribution over substitution models, you can use the bModelTestAnalyser. Start the AppStore application, and launch bModelTestAnalyser
A dialog pops up where you can specify the trace log file and set some parameters:
After you click OK, a window pops up listing some statistics:
and after a while, your web-browser should show a new page (one for each partition that uses bModelTest) that looks something like this:
For problems with low ESSs and setting up, see the trouble shooting page.
A more detailed description of the method can be found in Remco Bouckaert and Alexei Drummong. "bModelTest: Bayesian phylogenetic site model averaging and model comparison", BioArxiv, 2015. http://www.biorxiv.org/content/early/2015/06/11/020792