Danushka,

to be honest I am not so expert in this subject. I suggest you to ask to Enrico Ercole of the University of Turin (Italy). His email address is:

[email protected]

Write him telling that I have suggested you to ask him.

Regards,

Giampaolo

Dear Giampaolo,

Many thanks for the suggestion. I will contact him.

Best Regards

Dhanu 

Hello Dhanushka,

You should run ModelTest to test what model is best for your data. The best model varies based on your data so there is no one model that fits all data sets. ModelTest is also on CIPRES and pretty easy to run. Hope this helps.

Best Regards

Dear Dhanushka,

You can simply follow thses steps-

Create a new folder

Enter Folder Details

Save

Click on your new folder

Upload data (.phy)

Create new task

Select input data from uploaded data

Select Tool "jModelTest2 on XSEDE (2.1.6)"

Set parameters

Clustering search (-H): BIC

Base tree topology for likelihood calculations. (-t): Here I prefer to use Maximum Likelihood; however, you could use other options like- Fixed BIONJ  OR Neighbor-Joining.

Select Information Criterion (choose all needed): Bayesian Information Criterion

Do not change other parameters

Save parameters

Add a Description of your Work.

Save and Run task

Wait until Action column display as View Output

Meanwhile click Refresh Tasks few times

"PROCESS_OUTPUT STDOUT" will provide you the BEST MODEL for your alignment.

Click on View option.

The model with the LOWEST BIC score is the Best Fit Model for your alignment.

Hope this information helps.

Thanks 

Arun

Dear Arun,

Many thanks for the detailed procedure. I will follow you.

Best Regards

Dhanu