[gmx-users] grompp warnings/errors I haven't seen before

[David Mobley]

All,

I'm trying to simulate a peptide/water/cosolvent mix in gromacs
(3.3.2). I've set up a topology and coordinates and everything seems
fine until I run grompp, at which point I get

WARNING 1 [file "solvated.top", line 17]:
  Overriding atomtype HO
WARNING 2 [file "solvated.top", line 22]:
  Overriding atomtype OH
WARNING 3 [file "solvated.top", line 26]:
  Overriding atomtype H1
Generated 210 of the 210 non-bonded parameter combinations
Generating 1-4 interactions: fudge = 0.5
Generated 210 of the 210 1-4 parameter combinations
Excluding 3 bonded neighbours for solute 1
turning H bonds into constraints...
Excluding 1 bonded neighbours for SOL 4816
turning H bonds into constraints...
Excluding 3 bonded neighbours for TFE 805
turning H bonds into constraints...
processing coordinates...
double-checking input for internal consistency...
renumbering atomtypes...
converting bonded parameters...
#      BONDS:   4234
#     ANGLES:   11205
#     RBDIHS:   10605
#       LJ14:   10714
#     CONSTR:   2615
#     SETTLE:   4816
Walking down the molecule graph to make shake-blocks
Double sids (903, 904) for atom 22086
Double sids (903, 904) for atom 22087
Double sids (903, 904) for atom 22088
Double sids (903, 904) for atom 22089
Double sids (903, 904) for atom 22090
Double sids (903, 904) for atom 22091
Double sids (903, 904) for atom 22092
Assertion failed for "n == natoms" in file splitter.c, line 668
dump core ? (y/n):n

I've never seen this stuff before. First off, I don't understand the
warnings -- what's overriding those atom types? I don't see anything
wrong with them or anythign that should override them.

Second, what does the business about "Double sids" and "shake blocks"
mean? Clearly I've got something wrong in my topology or in the
relationship between my topology and gro file, but this is not at all
transparent.

I'll poke around in the source and see if I can figure out what's
going on but I thought I might be able to expedite things by an e-mail
to the list.

Thanks,
David Mobley
http://www.dillgroup.ucsf.edu/~dmobley