What to specify for substitution rate (mutation rate) and clock.date/ucld?

[Ch. Julian]

Hi everyone,

I have 200 viral time-stamped sequences (one gene, time as days) and a reported rate of evolution for this gene (6.02e-4 subs/site/year).

Back, in Beauti 1.8, when using other time-stamped datasets (time as years), I placed that prior information on the ucld.mean (a prior exponential distribution with a mean 6.02e-4).

In Beauti 2.1.1, I am a bit confused about what goes into the “substitution rate (mutation rate)” from both the Site Model and the Priors panels (The default is a prior 1/X distribution) . Is different from the information that I should enter into the clock.rate and the ucldMean boxes? Am I missing something?

Best regards

[Remco Bouckaert]

Hi Julian,

The substitution rate should probably be left at 1.0 -- it is best used for determining relative rates among partitions, which is what you get when the fixed-mean-mutation-rate check-box is selected. Otherwise the prior should go on the clock rate.

If you have a single gene, do not estimate both substitution rate and clock rate, since they will be unidentifyable. Just keep the substitution rate fixed at one, and do not select the estimate check-box. No prior on substitution rate will appear in the priors panel if you keep it this way.

BEAUti will automatically estimate the clock rate if you have a calibration or tip dates on your samples. When the clock rate is estimated, a prior on the clock rate appears in the priors panel, so that is where you specify the exponential prior.

Hope this helps,

Remco