MrBayes on XSEDE repeatedly stoping after 331000 generations

[jhorn]

Hi,

I am attempting to run a MrBayes 3.2.7a analysis (on XSEDE) and wish to run 2 run (with 4 chains each) for 40 million generations. I set the maximum hours to run for 40. However, the analysis terminates in less than 10 minutes after 331000 generations. I've attempted to run this analysis three times, always with the same result.

I've purchased plenty of CPU time, so I don't think that is an issue with respect to the premature termination of the run.

What advice does this group have on overcoming the issue?

 Best,

Jay

[Mark Miller]

Hi Jay,

Your runs stopped because they reached the configured stoprun value:

In the parameters, it looks like this:
stopruleval_ stoprule=Yes
stopval_  0.01

You need to make the stopval_ smaller if you want to have it continue, or disable the stopruleval parameter.

Here is how it looks in the interface:

Stop early if the convergence diagnostic falls below the stop value? + Yes No
Please enter the stop value

You can turn it off by clicking "no" to the first one. Then it will run until you reach your prescribed generations.

Or enter a value less than 0.01 on the second one, and it will stop when it reaches the specified value..

Let me know if you have further problems.

Mark