[gmx-developers] FEP Instability problem

[David van der Spoel]

Dear all,

I'm trying to FEP away a TIP4P water molecule from a water box in gmx 
4.5.3, but getting a jump in dH/dlambda after some time (around 100 ps), 
from around -5 to -600 kJ/mol/lambda. I'm using softcore with the 
following parameters:

sc-alpha = 0.5
sc-power = 1
sc-sigma = 0.3
delta-lambda             = 0
init-lambda = 0.1
couple-lambda0 = vdw
couple-lambda1 = vdw-q
free-energy              = yes

I've also tried using
sc-alpha = 0
sc-power = 0

Any clues what could go wrong?

Could it be due to the virtual sites?

Although not strictly a developer problem I took the liberty of sending 
this to the developer list anyhow, under the rationale that if a 
developer has problems with simulations, others may have them too, and 
something might need fixing.

-- 
David van der Spoel, Ph.D., Professor of Biology
Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone:	+46184714205.
spoel at xray.bmc.uu.se    http://folding.bmc.uu.se