[ESPResSo] Rigid constraints anyone?

[Aimee Bailey]

Greetings,

My name is Aimee Bailey (Imperial College London).  I developed a
couple algorithms for speedy constraint calculations - much faster
than SHAKE/RATTLE - and was wondering if anyone using espresso is
using rigid constraints? or would like to?

The algorithms aren't completely general - they exploit the geometry
of the system.  The first one (MILC SHAKE) exploits a linear geometry
and would be applicable to coarsegrained models (a factor of 3 faster
for n=10, up to two to three orders of magnitude for very large
systems, n>100).

The second one (MILCH SHAKE) is a more general method that works very
well for more branched geometries.  I tested it on linear alkanes with
explicit hydrogen.  It was 2 times faster than SHAKE for ethane up to
an order of magnitude faster for dodecane and higher molecular
weights.  I have preprints of papers that I can send to anyone who is
interested.

I just wanted to check to see if there is a need for them in the
espresso community.  One additional advantage is that they work well
for ringed systems, where SHAKE/RATTLE break down. I'm currently
testing MILCH on a coarsegrained bile salt, for instance, with a
4-membered ring, and it is robust.

If anyone is interested, please send me an e-mail.  Thanks!

Best, Aimee
address@hidden
address@hidden