Toy dataset for Bayesian dating with MCMCtree

[bheimbu]

Hi,

would it be possible to provide a toy dataset for Bayesian dating with MCMCtree. I've followed your example commands and realized that the resulting *.ctl includes some mistakes, like this:

BDparas = 1,1,0.5    * birth rate, death rate, sampling priors for sampling times
# must not contain commas 
BDparas = 1 1 0.5    * birth rate, death rate, sampling priors for sampling times

Additonally, the usage of RootAge = <1.0 * safe constraint on root age, used if no fossil for root. is discouraged in later versions of PAML (latest: PAML v4.10). See here. I've excluded this line from the *.ctl file and used an upper bond instead in the *.rooted.nwk treefile ('U(1.0)').

However, I've tried with a small dataset of primates and always get Floating point exception (core dumped). Find attached all necessary files to reproduce the run.

mcmctree_primates.zip

Short update:

It works with another version of paml from here and some minor changes to the *.ctl file as suggested on your website:

The ckpfile and hessianfile options are new and only work with our modified PAML code. If you use another MCMCtree version/release, you can simply remove those options from control file and rename the hessian file to in.BV to run MCMCtree without any errors.

In addition to this, you have to include print = 1 * 0: no mcmc sample; 1: everything except branch rates 2: everything in the *.ctl file, otherwise the mcmcfile (MCMC log of parameters that can be examined in Tracer) will not be created.

Cheers Bastian

[bqminh]

Thanks for the comment, @piyumalanthony Can you have a look?

[piyumalanthony]

Thanks @bheimbu for reporting this issue. I'm currently checking on it and will get back to you as soon as possible.