[gmx-users] Polarized water model

[xieyh at hkusua.hku.hk]

Dear users:

I would like to use polarized water model (shell MD) in my simulations of
protein solvated in SW water.

According to the instruction in the mailing list. I go on as the follows.

1. Using "mkyaw" program to transform "spc216.gro" box to "sw216.gro", as well
as using "sw.itp" instead of "spc.itp". 

But has this SW model been already energy minimized? Or, I need a separate
energy minization before I add my protein into this solvent?

2. When energy minimization, as mentioned in GMX manual, emtol should be 1
KJ/(mol.nm) in the case of shell MD, is it possible? Even I have only SW water
in my system, such small value is hard to obtain. If I set e.g. emtol = 1000,
the errors in PR step will be emerged (as the following). Can anybody tell me
how to deal with such problem? 

3. Also, does EM for shell MD run in double precision? If yes, how to realize
it?


PR:
....
MDStep=   58/ 0 lamb:      1, EPot: -1.03722727e+05
MDStep=   58/ 1 lamb:      1, EPot: 6.19446875e+05, rmsF: 6.19815250e+05
MDStep=   58/ 2 lamb:    0.8, EPot: 1.81001281e+05, rmsF: 1.57898906e+04
MDStep=   58/ 3 lamb:   0.64, EPot: 7.93818125e+04, rmsF: 1.18168262e+04
MDStep=   58/ 4 lamb:  0.512, EPot: 9.21008375e+05, rmsF: 2.28073525e+06
MDStep=   58/ 5 lamb: 0.4096, EPot: 1.16089463e+04, rmsF: 5.79389609e+04
MDStep=   58/ 6 lamb: 0.32768, EPot: -3.77739141e+04, rmsF: 5.81182715e+03
MDStep=   58/ 7 lamb:      1, EPot: -2.15094688e+04, rmsF: 6.25664941e+03
MDStep=   58/ 8 lamb:    0.8, EPot: -5.09686641e+04, rmsF: 6.39271191e+03
MDStep=   58/ 9 lamb:   0.64, EPot: -7.30289453e+04, rmsF: 9.48346387e+03
MDStep=   58/10 lamb:  0.512, EPot: -6.75546953e+04, rmsF: 7.49182617e+03
MDStep=   58/11 lamb: 0.4096, EPot: -8.74430703e+04, rmsF: 2.01819172e+05
MDStep=   58/12 lamb: 0.32768, EPot: -6.31760078e+04, rmsF: 1.07763262e+04
MDStep=   58/13 lamb: 0.262144, EPot: -4.65006758e+04, rmsF: 4.00746558e+03
MDStep=   58/14 lamb:      1, EPot: -1.04746039e+05, rmsF: 1.01801211e+04
MDStep=   58/15 lamb:    0.8, EPot: -9.85891797e+04, rmsF: 6.93954883e+03
MDStep=   58/16 lamb:   0.64, EPot: -1.10889062e+05, rmsF: 1.17012334e+04
MDStep=   58/17 lamb:  0.512, EPot: -1.01790453e+05, rmsF: 1.59684484e+05
MDStep=   58/18 lamb: 0.4096, EPot: -9.22433828e+04, rmsF: 4.88364941e+03
MDStep=   58/19 lamb: 0.32768, EPot: -8.15981172e+04, rmsF: 3.47391772e+03
EM did not converge in 20 steps
....
segmentation fault


Many thanks.


Xie YH
Hong Kong University