[R] Getting Started with Bayesian MCMC

[Ben]

Hi all,

I would like to start to use R's MCMC abilities to compute answers in
Bayesian statistics.  I don't have any specific problems in mind yet,
but I would like to be able to compute/sample posterior probabilities
for low-dimensional custom models, as well as handle "standard"
Bayesian cases like linear regression and hierarchical models.

R clearly has a lot of abilities in this area:

    http://cran.r-project.org/web/views/Bayesian.html

--enough to be confusing!  For instance, there are apparently three
separate interfaces to JAGS, and I'm not even sure I want/need to
interface to JAGS at all.

Can someone please get me started?  Are there a handful of "must-have"
packages and software that everyone (who uses MCMC regularly) uses?

Any responses are appreciated,

-- 
Ben