Testing a force topology in mrBayes

Question

[Dr Reema Singh]

Hello Everybody,

I have a question related to "Testing a topological hypothesis". I have tried to run the same procedure as mentioned in the tutorial in order to check whether two species form monophyletic group or not. As far as I understand the manual- - by setting hard constraint we are forcing two species to form a monophyletic group every time. And by setting the negative constraint we should reproduce the original topology. However, in my case , after running the analysis as mentioned in the manual- both hard and negative constraint are producing the same tree. After running the stepping stone analysis, I am getting some weird output in log file:-

24600 -- [nan] (nan) (nan) (nan) * [nan] (nan) (nan) (nan) (...0 remote chains

 Run   Marginal likelihood (ln)
       ------------------------------
         1          nan
         2          nan
       ------------------------------
       Mean:        nan

As I am running the entire analysis on the cluster. So, here's the command I am using for this analysis:-

begin mrbayes;
   set autoclose=yes nowarn=yes;
   execute Set1_Out_Ehist_Concat.nexus;
   lset rates=invgamma;
   constraint subell = Asub Aell;
   constraint nosubell = Asub Aell;
   prset topologypr = constraints(subell);
   mcmc ngen=100000 samplefreq=100 printfreq=100 diagnfreq=1000
file=PositiveConstraint;
   sump burnin=25000;
   sumt burnin=25000;
   prset topologypr = constraints(nosubell);
   mcmc ngen=100000 samplefreq=100 printfreq=100 diagnfreq=1000
file=NegativeConstraint;
   sump burnin=25000;
   sumt burnin=25000;
   prset topologypr = constraints(subell);
   ss ngen=250000 diagnfreq=2500 filename=positive;
   prset topologypr = constraints(nosubell);
   ss filename=negative;
   quit;
  end;

Here the Asub and Aell are the two species name that I want to be force to form a monophyletic group in the hard constraint.

Now I don't understand what is happening here. And my questions are :-

1. Why I am getting the same tree after setting hard and negative constraint?
2. What does this "nan" mean in the stepping stone analysis?
3. How do I known if the procedure I am running is correct or not?

As these question making me very confuse now. So I would highly appreciate any help and clarification on this questions. I have also posted this question on the mrbayes mailing list (http://sourceforge.net/p/mrbayes/mailman/message/34174009/), but I didn't get any reply.

Looking forward for the answer.

Many Thanks,
Reema Singh

Answer 1

[17th Level Botanist]

For Q1:

It looks to me like you set the exact same constraint twice. I think you need to revise

constraint nosubell = Asub Aell;

to

constraint nosubell negative = Asub Aell;

The default constraint is hard, which is why you do not need to specify that one.