Hello all,
I've pushed a software project that I've been working on up as a branch,
called amber_emil. The project acronym is EMIL (not sure what is stands
for, Einstein Molecule In Liquid, maybe).
The purpose of the code is to do an absolute free energy calculation, by
following these steps:
1) Define an analytically tractable Hamiltonian which has equilibrium
conformations somewhat similar to the start config of the simulation
2) Mix forces every timestep (either PME or GBSA) with the EMIL Hamiltonian
according to some lambda on [0,1].
... so when the runs are finished, you can do a thermodynamic integration,
to find delta free energy between the tractable model (where the free
energy can be output trivially) and the AMBER Hamiltonian.
The organisation of the code is as a library, compiled from a load of C++
code in $AMBERHOME/src/emil. It plugs into runmd.F90 currently, with two
function calls "emil_init()" and "emil_step()". I've used autotools to
manage the build of lthe library, although I have to admit that it is my
first time with them so there may be schoolboy errors which others can
point out.
I've added a section in Amber12.lyx, following on from the chapter on TI,
and some test scripts (emil_gbsa and emil_pmemd). I've verified the tests
for (intel+gnu)x(serial+mpi) on intel CPUs only.
The paper that goes with the code is open-access on JCTC:
http://pubs.acs.org/doi/full/10.1021/ct3005968
I'm coming to the meeting in Utah, so there should be plenty of time to
discuss, but if you do try it out then please let me know what you think as
you go along, perspectives from others who are working on TI-type stuff
should be of particular interest.
Thomas and others working in Yurp: I am in Luxembourg, so visits to
neighbouring countries to give a seminar or something are no problem if you
would like to discuss offline.
Josh Berryman
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Fri Feb 01 2013 - 02:30:02 PST
