basic question from a new BEAST user

161 views
Skip to first unread message

oizo

unread,
Mar 25, 2010, 7:44:18 AM3/25/10
to beast-users
Hi all,

I have a basic question:
I ran an anlysis on some mtDNA sequences data with beast. For the
clock model I specified an uncorrelated lognormal relaxed clock and
ticked the box "estimate"; and i specified a normally distributed mean
rate of evolution over the whole tree of partition of 1.89 ± 0.35% per
million year. I had no fossil calibration to run this analysis.
My question is : when looking at the results in Tracer, what unit are
the tmrca mean expressed in? and the ESS?

I'm a bit lost and confused at the moment (maybe because I am just
starting to use these software....).

Thank you for any help/comments/advices.


julien

unread,
Mar 31, 2010, 5:37:10 PM3/31/10
to beast-users
If you set the value 0.0189 correctly, then edge lengths are given in
millions years... You might also use a truncated normal prior, not to
sample rates below 0 nor above 1... which biologically makes senses.
Theoritician much better than me alternatively suggest using log-
normal priors, see here:

http://groups.google.com/group/beast-users/browse_thread/thread/3ff546232471c381/31d45bc93de9d945?lnk=gst&q=fix+mean+subsitution+rate#31d45bc93de9d945

ESS have no unit (at least I think)... They just give you an idea how
well the Markov Chain is mixing (=sampling the distribution)... The
greater values are (should be >>200), the better the chain mixes.

Good luck

Julien

Reply all
Reply to author
Forward
Message has been deleted
0 new messages