[gmx-users] g_bar and second law violations

[Sanku M]

Hi,
  I am trying to compute the free energy change using g_bar: I am using two-stage decoupling process as in Justin Lemkul's tutorial.
But , I get following warning in all cases regarding violation of second law of thermodynamics. Can someone explain what is this error about and which lambda I should look for?
Here is the output.

 Detailed results in kT (see help for explanation):

 lam_A  lam_B      DG   +/-     s_A   +/-     s_B   +/-   stdev   +/- 
 0.000  0.100   -0.39  0.08   -0.64  0.06   -0.09  0.20    0.88  0.05
 0.100  0.150    0.08  0.04    0.07  0.03    0.09  0.05    0.43  0.01
 0.150  0.200    0.12  0.02   -0.11  0.03    0.36  0.06    0.54  0.02
 0.200  0.300    0.54  0.06   -0.05  0.07    1.25  0.15    1.24  0.03
 0.300  0.350    0.21  0.03    0.15  0.04    0.31  0.06    0.72  0.01
 0.350  0.400   -0.01  0.02    0.11  0.03   -0.08  0.04    0.64  0.01
 0.400  0.500   -0.37  0.03    0.22  0.02   -0.78  0.06    1.07  0.01
 0.500  0.600   -0.58  0.01    0.17  0.04   -1.17  0.04    0.99  0.01
 0.600  0.700   -0.74  0.03    0.16  0.02   -1.51  0.07    0.89  0.01
 0.700  0.750   -0.37  0.01   -0.02  0.02   -0.69  0.03    0.47  0.01
 0.750  0.800   -0.26  0.01   -0.07  0.01   -0.42  0.02    0.53  0.01
 0.800  0.850   -0.19  0.01    0.03  0.01   -0.38  0.03    0.57  0.01
 0.850  0.900   -0.24  0.02    0.06  0.01   -0.50  0.04    0.58  0.01
 0.900  1.000   -0.47  0.07   -0.04  0.05   -0.12  0.34    1.13  0.06

WARNING: Some of these results violate the Second Law of Thermodynamics: 
         This is can be the result of severe undersampling, or (more likely)
         there is something wrong with the simulations.


Final results in kJ/mol:

lambda  0.000 -  0.100,   DG -0.97 +/-  0.19
lambda  0.100 -  0.150,   DG  0.20 +/-  0.09
lambda  0.150 -  0.200,   DG  0.30 +/-  0.05
lambda  0.200 -  0.300,   DG  1.34 +/-  0.15
lambda  0.300 -  0.350,   DG  0.52 +/-  0.07
lambda  0.350 -  0.400,   DG -0.02 +/-  0.05
lambda  0.400 -  0.500,   DG -0.92 +/-  0.08
lambda  0.500 -  0.600,   DG -1.45 +/-  0.03
lambda  0.600 -  0.700,   DG -1.86 +/-  0.08
lambda  0.700 -  0.750,   DG -0.91 +/-  0.03
lambda  0.750 -  0.800,   DG -0.64 +/-  0.03
lambda  0.800 -  0.850,   DG -0.48 +/-  0.03
lambda  0.850 -  0.900,   DG -0.61 +/-  0.04
lambda  0.900 -  1.000,   DG -1.17 +/-  0.18

total   0.000 -  1.000,   DG -6.68 +/-  0.30