Error: atoms lost

Dear users,

I am using the following commands in my input file and my data file is a box of water molecules only. I got an error indicating: " lost atoms: original 11763 current 11761 " could you please tell me where is the problem?
Thanks.
Neda

Intialization

units real
dimension 3
boundary p p p
atom_style full

Atom Definition

read_data data.mysys
replicate 1 1 1

#The SPC water potential
pair_style lj/cut/coul/cut 9.8 9.8
pair_coeff 2 2 0.15535 3.166
pair_coeff 1 * 0.00000 0.0000
bond_style harmonic
angle_style harmonic
dihedral_style none
improper_style none
bond_coeff 1 100.00 1.000
angle_coeff 1 100.00 109.47
special_bonds lj/coul 0.0 0.0 0.5

---------- Define neighbour list ---------------------

neighbor 3.0 bin
neigh_modify delay 10 check yes

---------- Dump Options ---------------------

dump 1 all xyz 1 dump.xyz

---------- Run Minimization ---------------------

reset_timestep 0
fix 1 all box/relax iso 0.0 vmax 0.001 #tri=triclinic pressure, vmx=max volume change
thermo 10
thermo_style custom step pe lx ly lz press
min_style cg
minimize 1e-25 1e-25 5000 10000

write_restart restart.relax

print “minization was already done!”

undump 1
unfix 1
#--------------- End of minimization section-------------

#----------- Equilibrium ---------------------------------

velocity all create 300 435628 dist gaussian mom yes rot yes
fix 2 all nvt temp 300 300 10 drag 1.0
reset_timestep 0
timestep 1
thermo 100
thermo_style custom step time timeremain temp press pe ke vol
dump 2 all xyz 100 dump2.xyz

run 10000
unfix 2

Dear users,
I am using the following commands in my input file and my data file is a
box of water molecules only. I got an error indicating: " lost atoms:
original 11763 current 11761 " could you please tell me where is the
problem?

​SPC is a rigid water potential, but i do not see you using fix shake or
fix rattle anywhere. for a flexible SPC model, however, your bond and angle
force constants are not correct (when using shake/rattle, they don't
matter) and your time step is too large.​

axel.