Definition of the rapid bootstrap

[Juan Carlos Zamora Señoret]

Hi all!

I would like to ask a question about the actual definition of the rapid bootstrap algorithm, since it was always to me a bit confusing to see the use of "rapid bootstrap" with a search under GTRGAMMA for computing the bootstrap trees. I know that this is a really basic question, but I have seen contrasting views in different places.

I always though that a rapid bootstrap uses the CAT approximation, as indicated in Stamatakis et al. (2008) that we all know, and that this was the difference between a normal bootstrap (computing replicates with GTRGAMMA) and a rapid bootstrap (computing replicates with GTRCAT). Is this intepretation correct? Is there any "rapid bootstrap" algorithm that uses GTRGAMMA, and then how is this different from a normal bootstrap (i.e., how is the speed increased if the algorithm is the same)?

Thank you very much!

 

Best regards,

Juan Carlos Zamora

[Alexandros Stamatakis]

Dear Juan,

> I would like to ask a question about the actual definition of the rapid
> bootstrap algorithm, since it was always to me a bit confusing to see
> the use of "rapid bootstrap" with a search under GTRGAMMA for computing
> the bootstrap trees. I know that this is a really basic question, but I
> have seen contrasting views in different places.
> I always though that a rapid bootstrap uses the CAT approximation, as
> indicated in Stamatakis et al. (2008) that we all know, and that this
> was the difference between a normal bootstrap (computing replicates with
> GTRGAMMA) and a rapid bootstrap (computing replicates with GTRCAT).

GTRGAMMA was added later to also work under rapid bootstrap, if you read
the paper from 2008 you will see that the rapid BS search algorithm is
also a bit more superficial than the standard more thorough RAxML search
algorithm.

> Is
> this intepretation correct? Is there any "rapid bootstrap" algorithm
> that uses GTRGAMMA,

Yes, see above.

> and then how is this different from a normal
> bootstrap (i.e., how is the speed increased if the algorithm is the same)?

Just read the 2008 paper please.

Alexis

>
> Thank you very much!
> Best regards,
>
> Juan Carlos Zamora
>

> --
> You received this message because you are subscribed to the Google
> Groups "raxml" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to raxml+un...@googlegroups.com
> <mailto:raxml+un...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.

--
Alexandros (Alexis) Stamatakis

Research Group Leader, Heidelberg Institute for Theoretical Studies
Full Professor, Dept. of Informatics, Karlsruhe Institute of Technology

www.exelixis-lab.org