[gmx-users] g_hbond results inconsistent between v4.0.7 and v4.5.1

[Robin C. Underwood]

Hi All:

Has there been further resolution on the difference between g_hbond in v 4.0 and
4.5 (I'm running 4.5.3)? For counting water-water (tip4p) h-bonds in a box of
506 waters I get the following nhbdist output, which appears to neglect pbc
since there are such a large number of unbound donor-H (and the other columns
are consequently too low):


@    title "Number of donor-H with N HBs"
@    xaxis  label "Time (ps)"
@    yaxis  label "N"
@TYPE xy
@ view 0.15, 0.15, 0.75, 0.85
@ legend on
@ legend box on
@ legend loctype view
@ legend 0.78, 0.8
@ legend length 2
@ s0 legend "0 HBs"
@ s1 legend "1 HB"
@ s2 legend "2 HBs"
@ s3 legend "3 HBs"
@ s4 legend "Total"
 0.00000e+00       706       174       132         0       438
 1.00000e-01       708       163       140         1       446
 2.00000e-01       707       168       137         0       442
 3.00000e-01       712       161       136         3       442
 4.00000e-01       713       165       134         0       433
 5.00000e-01       709       169       133         1       438
 6.00000e-01       705       169       137         1       446
 7.00000e-01       699       169       142         2       459
 8.00000e-01       710       167       133         2       439
 9.00000e-01       700       189       122         1       436
 1.00000e+00       707       169       136         0       441
 1.10000e+00       705       163       144         0       451
 1.20000e+00       707       171       133         1       440
 1.30000e+00       709       158       145         0       448
 1.40000e+00       718       155       138         1       434
 1.50000e+00       711       164       137         0       438
 1.60000e+00       713       160       139         0       438
 1.70000e+00       702       181       129         0       439
 1.80000e+00       707       162       142         1       449
 1.90000e+00       706       167       139         0       445
 2.00000e+00       709       168       135         0       438
 2.10000e+00       708       167       136         1       442
 2.20000e+00       708       171       133         0       437
 2.30000e+00       709       162       140         1       445
 2.40000e+00       698       171       142         1       458
 2.50000e+00       701       177       133         1       446
 2.60000e+00       706       177       129         0       435
 2.70000e+00       708       178       126         0       430
 2.80000e+00       698       187       127         0       441
 2.90000e+00       700       190       120         2       436
 3.00000e+00       703       176       131         2       444
 3.10000e+00       702       181       127         2       441
 3.20000e+00       694       184       134         0       452


I've tried playing with different index assignments, and I get what appears to
be a reasonable output for "0 HBs" by making the two groups appear different,
excluding the dummy atom from one. Although, because the program then double
counts, the other columns don't convey the intended information. 


Robin 






-- 
Robin C. Underwood
Chemistry Department
560 Oval Drive
West Lafayette, IN 47907