Thank you. I'm relieved to hear the moltemplate tutorials might be useful.
But I hope I did not raise your hopes too much. Moltemplate does not
automatically convert PDB files into coarse grained models. Also: I
confess moltemplate is not easy to use (just like lammps!). Usability
will improve over time.
It's okay to post a question (or email me a question) if you get stuck.
Cheers and thanks for the reply! (Boosts moral.)
Andrew
P.S.
As you mentioned, you still have to figure out the force-field
you want to use to solve your problem. Many people create a new one.
I suspect you will need some local guidance, or read the papers in
your field for inspiration. Nobody outside your area of research can
(or will be willing to) choose a force-field for you. Fortunately
LAMMPS offers a very large number of force-field styles to choose
from.
These days, methods like Inverse Monte-Carlo, Boltzmann
Inversion and force-matching are popular. (For software, try
http://code.google.com/p/magic or http://www.votca.org/) These
programs can help you generate a realistic force-field if you need
one. (However to use, them you have to collect data from an all-atom
simulation, or from the PDB library.)
Fortunately, sometimes you don't need it to be very realistic. (I
never was.) It depends on the question you are asking (and the
temperament of the professor you are working with).
Good luck.